Understanding structural isomerization during ruthenium-catalyzed olefin metathesis: A deuterium labeling study

Article abstract of DOI:10.1021/om0509894

A deuterium labeling study was undertaken to determine the mechanism of olefin isomerization during the metathesis reactions catalyzed by a second-generation Grubbs catalyst (2). The reaction of allyl-1,1-d₂ methyl ether with 2 at 35°C was followed by ¹H and ²H NMR spectroscopy. The evidence of deuterium incorporation at the C-2 position of the isomerized product, trans-propenyl methyl ether, led to the conclusion that a metal hydride addition - elimination mechanism was operating under these conditions. Consequently, complex 8, an analogue of 2 bearing deuterated o-methyl groups on the aromatic rings of the NHC ligand, was synthesized to investigate the role of the NHC ligand in the formation of hydride species. Thermal decomposition of benzylidene 8 and methylidene 8′ was monitored by ²H NMR spectroscopy; no deuteride complex was detected in either case. The decomposition mixtures were tested for isomerization activity with benchmark 1-octene but did not match the isomerization rates observed with 2 under similar metathesis conditions. Reaction of complex 8 with various olefmic substrates not only confirmed the formation of a deuteride complex but also revealed the existence of a competitive H/D exchange process between the CD ₃ groups on the NHC ligand and the C-H bonds of the substrate. We propose that the exchange is promoted by a ruthenium dihydride intermediate whose formation is closely related to the methylidene decomposition.

Full text of DOI:10.1021/om0509894

6074
Organometallics 2006, 25, 6074-6086
Understanding Structural Isomerization during
Ruthenium-Catalyzed Olefin Metathesis:
A Deuterium Labeling Study
Florence C. Courchay, John C. Sworen, Ion Ghiviriga, Khalil A. Abboud, and
Kenneth B. Wagener*
The George and Josephine Butler Polymer Research Laboratory, Department of Chemistry,
UniVersity of Florida, GainesVille, Florida 32611
ReceiVed NoVember 17, 2005
A deuterium labeling study was undertaken to determine the mechanism of olefin isomerization during
the metathesis reactions catalyzed by a second-generation Grubbs catalyst (2). The reaction of allyl-1,1-
1
2
2
d methyl ether with 2 at 35 °C was followed by H and H NMR spectroscopy. The evidence of deuterium
incorporation at the C-2 position of the isomerized product, trans-propenyl methyl ether, led to the
conclusion that a metal hydride addition-elimination mechanism was operating under these conditions.
Consequently, complex 8, an analogue of 2 bearing deuterated o-methyl groups on the aromatic rings of
the NHC ligand, was synthesized to investigate the role of the NHC ligand in the formation of hydride
2
species. Thermal decomposition of benzylidene 8 and methylidene 8′ was monitored by H NMR
spectroscopy; no deuteride complex was detected in either case. The decomposition mixtures were tested
for isomerization activity with benchmark 1-octene but did not match the isomerization rates observed
with 2 under similar metathesis conditions. Reaction of complex 8 with various olefinic substrates not
only confirmed the formation of a deuteride complex but also revealed the existence of a competitive
H/D exchange process between the CD groups on the NHC ligand and the C-H bonds of the substrate.
3
We propose that the exchange is promoted by a ruthenium dihydride intermediate whose formation is
closely related to the methylidene decomposition.
Introduction
The discovery of well-defined, functional-group-tolerant
ruthenium carbene catalysts such as Grubbs’ (PCy3)2(Cl)2Rud
1
CHPh (1) has considerably broadened the scope of olefin
metathesis.2 The versatility of this class of catalysts has
contributed to their growing popularity and motivated new
developments. A “second generation” of catalytic systems was
introduced as one N-heterocyclic carbene replaced a phosphine
ligand (2 and 3 in Figure 1), further enhancing their catalytic
Figure 1. Olefin metathesis catalysts.
3,4
modifications have been reported in order to tune the activity
and/or the stability of the catalyst according to specific reaction
conditions. For example, the exchange of a phosphine for an
isopropoxy-tethered benzylidene conferred an exceptional stabil-
ity to the catalyst in solution, which has allowed the catalyst to
be recycled after ring-closing metathesis (RCM) or cross-
activity and the thermal stability. Since then, many ligand
*
To whom correspondence should be addressed. E-mail: wagener@
chem.ufl.edu.
(
1) (a) Schwab, P.; Grubbs, R. H.; Ziller, J. J. Am. Chem. Soc. 1996,
1
18, 100-110. (b) Dias, E. L.; Nguyen, S. T.; Grubbs, R. H. J. Am. Chem.
Soc. 1997, 119, 3887-3897. (c) Ulman, M.; Grubbs, R. H. Organometallics
5
metathesis (CM) reactions. Also, addition of more labile
1
6
998, 17, 2484-2489. (d) Ulman, M.; Grubbs, R. H. J. Org. Chem. 1999,
4, 7202-7207.
ligands, such as pyridine derivatives, enabled fast initiation rates
particularly desirable for living ring-opening metathesis poly-
(2) For relevant reviews, see: (a) Handbook of Metathesis; Grubbs, R.
H., Ed.; Wiley-VCH: Weinheim, Germany, 2003. (b) Trnka, T. M.; Grubbs,
R. H. Acc. Chem. Res. 2001, 34, 18-29. (c) F u¨ rstner, A. Angew. Chem.,
Int. Ed. 2000, 39, 3012-3043. (d) Grubbs, R. H.; Chang, S. Tetrahedron
6
merization (ROMP).
However, in addition to their high reactivity in metathesis
chemistry, these ruthenium catalysts have been shown to
undergo non-metathesis transformations. Among them, alkene
1
998, 54, 4413-4450. (e) Schuster, M.; Blechert, S. Angew. Chem., Int.
Ed. 1997, 36, 2036-2056. (f) Armstrong, S. K. J. Chem. Soc., Perkin Trans.
1
4
7
1998, 371-388. (g) Frenzel, U.; Nuyken, O. J. Polym. Sci.: Chem. 2002,
0, 2895-2916.
isomerization has emerged as an important side reaction of
(3) (a) Scholl, M.; Ding, S.; Lee, C.; Grubbs, R. H. Org. Lett. 1999, 1,
9
53-956. (b) Sanford, M. S.; Ulman, M.; Grubbs, R. H. J. Am. Chem.
(5) (a) Kingsbury, J. S.; Harrity, J. P. A.; Bonitatebus, P. J.; Hoveyda,
A. H. J. Am. Chem. Soc. 1999, 121, 791-799. (b) Garber, S. B.; Kingsbury,
J. S.; Gray, B. L.; Hoveyda, A. H. J. Am. Chem. Soc. 2000, 122, 8168-
8179. (c) For a review on Ru-bearing bidentate carbenes: Hoveyda, A. H.;
Gillingham, D. G.; Van Veldhuizen, J. J.; Kataoka, O.; Garber, S. B.;
Kingsbury, J. S.; Harrity, J. P. A. Org. Biomol. Chem. 2004, 2, 8-23.
(6) (a) Love, J. A.; Morgan, J. P.; Trnka, T. M.; Grubbs, R. H. Angew.
Chem., Int. Ed. 2002, 41, 4035-4037. (b) Love, J. A.; Sanford, M. S.;
Day, M. W.; Grubbs, R. H. J. Am. Chem. Soc. 2003, 125, 10103-10109.
Soc. 2001, 123, 749-750. (c) Sanford, M. S.; Love, J. A.; Grubbs, R. H.
J. Am. Chem. Soc. 2001, 123, 6543-6554.
(
4) (a) Scholl, M.; Trnka, T. M.; Morgan, J. P.; Grubbs, R. H.
Tetrahedron Lett. 1999, 40, 2247-2250. (b) Weskamp, T.; Kohl, F. J.;
Herrmann, W. A. J. Organomet. Chem. 1999, 582, 362-365. (c) Acker-
mann, L.; F u¨ rstner, A.; Westcamp, T.; Kohl, F. J.; Herrmann, W. A.
Tetrahedron Lett. 1999, 40, 4787-4790. (d) Huang, J.; Stephens, E. D.;
Nolan, S. P.; Pedersen, J. L. J. Am. Chem. Soc. 1999, 121, 2674-2678.
1
0.1021/om0509894 CCC: $33.50 © 2006 American Chemical Society
Publication on Web 11/23/2006

Ruthenium-Catalyzed Olefin Metathesis
Organometallics, Vol. 25, No. 26, 2006 6075
Figure 2. Decomposition products of ruthenium carbene complexes.
ruthenium-catalyzed metathesis, disturbing the product’s mi-
crostructure by apparent migration of the double bond along
Apart from standard hydrogenolysis,15 numerous pathways
have been proposed for the transformation of ruthenium carbene
into ruthenium hydride complexes. For example, the carbonyl
hydrides 4 and 5 (Figure 2), first isolated during the decomposi-
8
the alkyl chain. First observed on substrates containing allylic
9
functionalities in combination with first-generation catalysts,
1
6
double-bond isomerization has since been reported with second-
generation catalysts on a broad variety of substrates competi-
tively, sometimes prior to olefin metathesis.10,11Promoted by
diverse transition metals, olefin isomerization has been known
tion of alkoxycarbene complexes, were identified as products
of the reaction of 1 and 2 with primary alcohols in the presence
17
of a base. Although the mechanism is still at a speculative
1
6a,17a
stage,
for isomerization, hydrogenation,
both complexes are known to be efficient catalysts
3
17,18
19
to operate through two distinctive pathways: either through η -
and hydrovinylation.
allyls or through an alkyl intermediate involving a long-lived
metal hydride.12 Mechanistically, the π-allyl involves a formal
intramolecular 1,3-H shift, while the metal hydride operates
through an intermolecular addition-elimination with an inher-
ently competitive 1,2-H shift. Consequently, isotopic labeling
experiments have often been used to probe the nature of the
mechanism.13 However, despite the increasing number of
studies, the ruthenium intermediate responsible for this undesir-
able reaction has not been identified yet. Although isomerization
is often attributed to the formation of a ruthenium hydride in
situ, as a decomposition product of the original carbene
More recently, Grubbs has isolated the dinuclear ruthenium
hydride 6 as one of the decomposition products of the
methylidene intermediate 2′, formed during a metathesis cycle.20
Other studies include the possibility of ruthenium insertion into
the C-H bond of one of the methyl groups present on the NHC
1
7d,21
ligand.
This type of activation is common for complexes
containing NHC ligands and has been witnessed during the
preparation of ruthenium carbene 2; however, it has not yielded
1
7d
the hydride complex 7.
Acyclic diene metathesis (ADMET) has been used recently
for the modeling of precise ethylene/R-olefin copolymers.22
Therefore controlling isomerization is crucial, since it results
in irregular microstructures, whether it occurs on the monomer
or on the product (Figure 3). Previously, we have reported the
metathesis vs isomerization activity of ruthenium carbene
14
catalyst, the possibility of a π-allyl mechanism cannot be ruled
1
0d
out.
(7) (a) For a review on non-metathetic behavior of Ru carbenes, see:
Alcaide, B.; Almendros, P. Chem. Eur. J. 2003, 9, 1258-1262. (b) Schmidt,
B.; Pohler, M.; Costisella, B. J. Org. Chem. 2004, 69, 1421-1424. (c)
Faulkner, J.; Edlin, C. D.; Fengas, D.; Preece, I.; Quayle, P.; Richards, S.
N. Tetrahedron Lett. 2005, 46, 2381-2385. (d) Edlin, C. D.; Faulkner, J.;
Fengas, D.; Knight, C. K.; Parker, J.; Preece, I.; Quayle, P.; Richards, S.
N. Synlett 2005, 4, 572-577. (e) Hoye, T. R.; Zhao, H. Org. Lett. 1999, 1,
1
1
catalysts using the model substrate 1-octene. Even though
(14) For relevant reviews on the reactivity of ruthenium hydride
complexes, see: (a) Naota, T.; Takaya, H.; Murahashi, S. Chem. ReV. 1998,
98, 2599-2660. (b) Trost, B. M.; Toste, F. D.; Pinkerton, A. B. Chem.
ReV. 2001, 101, 2067-2096.
(15) (a) Drouin, S. D.; Yap, G. P. A.; Fogg, D. E. Inorg. Chem. 2000,
39, 5412-5414. (b) Fogg, D. E.; Drouin, S. D.; Zamanian, F. U.S. Pat.
Appl. 60/192394, priority date March 26, 2000. (c) Beatty, R. P.; Paciello,
R. A. U.S. Patent 5,554,778, 1996.
1
3
69-171. (f) Gurjar, M. K.; Yakambram, P. Tetrahedron Lett. 2001, 42,
633-3636.
(
8) For a short review on the isomerization behavior of ruthenium
carbenes, see: Schmidt, B. Eur. J. Org. Chem. 2004, 9, 1865-1880.
(9) (a) Miller, S. J.; Blackwell, H. E.; Grubbs, R. H. J. Am. Chem. Soc.
1
996, 118, 9606-9614. (b) Maynard, H. D.; Grubbs, R. H. Tetrahedron
(16) (a) Wu, Z.; Nguyen, S. T.; Grubbs, R. H.; Ziller, J. W. J. Am. Chem.
Soc. 1995, 117, 5503-5511. (b) Louie, J.; Grubbs, R. H. Organometallics
2002, 21, 2153-2164.
(17) (a) Dinger, M. B.; Mol, J. C. Organometallics 2003, 22, 1089-
1095. (b) Dinger, M. B.; Mol, J. C. Eur. J. Inorg. Chem. 2003, 2827-
2833. (c) Banti, D.; Mol, J. C. J. Organomet. Chem. 2004, 689, 3113-
3116. (d) Trnka, T. M.; Morgan, J. P.; Sanford, M. S.; Wilhem, T. E.; Scholl,
M.; Choi, T.-L.; Ding, S.; Day, M. D.; Grubbs, R. H. J. Am. Chem. Soc.
2003, 125, 2546-2558.
(18) (a) Yi, C. S.; Lee, D. W. Organometallics 1999, 18, 5152-5156.
(b) Yi, C. S.; Lee, D. W.; He, Z.; Rheingold, A. L.; Lam, K.-C.; Concolino,
T. E. Organometallics 2000, 19, 2909-2915. (c) Lee, H. M.; Smith, D.
C.; Zhengjie, H.; Stevens, E. D.; Yi, C. S.; Nolan, S. P. Organometallics
2001, 20, 794-797. (d) Dharmasena, U. L.; Foucault, H. M.; dos Santos,
E. N.; Fogg, D. E.; Nolan, S. P. Organometallics 2005, 24, 1056-1058.
(19) Esteruelas, M. A.; Werner, H. J. Organomet. Chem. 1986, 303, 221-
231.
Lett. 1999, 40, 4137-4140. (c) Bourgeois, D.; Pancrazi, A.; Ricard, L.;
Prunet, J. Angew. Chem. 2000, 112, 742-744; Angew. Chem., Int. Ed. 2000,
3
9, 725-728. (d) Edwards, S. D.; Lewis, T.; Taylor, R. J. K. Tetrahedron
Lett. 1999, 40, 4267-4270. (e) Schmidt, B.; Wildemann, H. J. Org. Chem.
000, 65, 5817-5822.
10) (a) F u¨ rstner, A.; Thiel, O. R.; Ackermann, L.; Schanz, H.-J.; Nolan,
2
(
S. P. J. Org. Chem. 2000, 65, 2204-2207. (b) Cadot, C.; Dalko, P. I.;
Cossy, J. Tetrahedron Lett. 2002, 43, 1839-1841. (c) Sutton, A. E.; Seigal,
B. A.; Finnegan, D. F.; Snapper, M. L. J. Am. Chem. Soc. 2002, 124,
1
3390-13391. (d) Bourgeois, D.; Pancrazi, A.; Nolan, S. P.; Prunet, J. J.
Organomet. Chem. 2002, 643-644, 247-252. (e) Sworen, J. C.; Pawlow,
J. H.; Case, W.; Lever, J.; Wagener, K. B. J. Mol. Catal. 2002, 194, 69-
7
8.
(11) (a) Lehman, S. E.; Schwendeman, J. E.; O’Donnell, P. M.; Wagener,
K. B. Inorg. Chim. Acta 2002, 345, 190-198. (b) Courchay, F. C.; Sworen,
J. C.; Wagener, K. B. Macromolecules 2003, 36, 8231-8239.
(12) Crabtree, R. H. The Organometallic Chemistry of the Transition
(20) Hong, S. H.; Day, M. W.; Grubbs, R. H. J. Am. Chem. Soc. 2004,
126, 7414-7415.
Metals, 2nd ed.; Wiley: New York, 1994.
13) (a) Hendrix, W. T.; Cowherd, F. G.; von Rosenberg, J. L. Chem.
Commun. 1968, 97-99. (b) Bergens, S. H.; Bosnich, B. J. Am. Chem. Soc.
(
(21) Jazzar, R. F. R.; Macgregor, S. A.; Mahon, M. F.; Richards, S. P.;
Whittlesey, M. K. J. Am. Chem. Soc. 2002, 124, 4944-4945.
(22) (a) Wagener, K. B.; Valenti, D. J.; Hahn, S. F. Macromolecules
1997, 30, 6688-6690. (b) Smith, J. A.; Brzezinska, K. R.; Valenti, D. J.;
Wagener, K. B. Macromolecules 2000, 33, 3781-3794. (c) Sworen, J. C.;
Smith, J. A.; Berg, J. M.; Wagener, K. W. J. Am. Chem. Soc. 2004, 126,
11238-11246.
1
1
991, 113, 958-967. (c) McGrath, D. V.; Grubbs, R. H. Organometallics
994, 13, 224-235. (d) Baudry, D.; Ephritikhine, M.; Felkin, H. J. Chem.
Soc., Chem. Commun. 1978, 694-695. (e) Clark, H. C.; Kurosawa, H. Inorg.
Chem. 1973, 12, 1566-1569. (f) Casey, C. P.; Cyr, C. R. J. Am. Chem.
Soc. 1973, 95, 2248-2253.

6076 Organometallics, Vol. 25, No. 26, 2006
Courchay et al.
Figure 4. Deuterium shift with a π-allyl hydride mechanism.
Figure 3. Effect of isomerization during ADMET polymerization.
tions of the metal hydride across the double bond results in a
1
3e
characteristic 1,2-H shift.
When a primary metal alkyl
metathesis consistently occurred more quickly, isomerization
products could represent as much as 80% of the reaction mixture
with catalysts containing NHC ligands. When other modified
undergoes a â-elimination of a different hydrogen, the resulting
olefin can reinsert in the metal-hydride bond to form a
secondary metal alkyl. Subsequent â-elimination of the ap-
propriate hydrogen results in an apparent 1,2-H shift, as
illustrated in Figure 5 for a deuterated olefin. A formal 1,3-H
shift can also be observed through the formation of a secondary
metal-hydride.
complexes were used, we observed that the isomerization
activity was closely related to the metathesis activity.23 Intrigued
by the apparent relationship between isomerization and metath-
esis behavior, we decided to investigate the mechanism of
isomerization under metathesis conditions. The reaction of allyl-
1
3a-c
1
Substrates such as allyl-1,1-d alcohol
sponding methyl ether
and the corre-
1
,1-d2 methyl ether with catalyst 2 was monitored by H and
2
1
3c-e
2
have often been used to probe the
H NMR spectroscopy to examine deuterium scrambling and
therefore determine the operating isomerization mechanism. We
also designed a deuterated version of catalyst 2 (complex 8 in
Scheme 1) to further demonstrate the role of the NHC ligand
in catalyst decomposition and assess the possibility of C-H
activation, which could be responsible for the formation of a
nature of the hydrogen shift in metal-catalyzed olefin isomer-
ization, particularly because of the irreversibility of the reaction.
Indeed, allyl-1,1-d2 alcohol will isomerize exclusively into
propionaldehyde-1,3-d2 in the case of a π-allyl hydride mech-
anism, while a mixture of partially deuterated propionaldehydes
with deuterium incorporation at the C-2 position will be expected
in the case of the metal hydride mechanism. Actually this
method is often taken as evidence for the π-allyl hydride
mechanism without investigating the inter-/intramolecularity of
the process.
2
metal hydride complex. Again, we used H NMR to monitor
the decomposition of complex 8 and observe its behavior in
the presence of olefinic substrates, looking for any deuterium
incorporation that would confirm the presence of a ruthenium
deuteride complex.
In the present case, allyl alcohol has been known to react
with ruthenium carbenes and form hydride complexes.7
Therefore, we have worked with allyl methyl ether only, which
in the end should provide us with the same set of observations.
A 2.3 M solution of allyl-1,1-d2 methyl ether (9) in C6D6 was
charged in an airtight NMR tube with 0.2 mol % of complex 2
and heated at 35 °C for 3 days to reproduce the conditions used
d,e,24
Results and Discussion
Mechanistic Studies. Even though both the π-allyl hydride
and the metal hydride addition-elimination mechanism result
in the same product, isotopic labeling studies can highlight their
differences. Indeed, the two pathways can be distinguished by
looking at the hydrogen shift undergone during the isomerization
reaction. The π-allyl mechanism, involving a metal with two
empty coordination sites that does not contain a hydride ligand,
is typically intramolecular. A free olefin coordinates to the metal
and undergoes an oxidative addition of the allylic C-H bond
to form a π-allyl metal hydride intermediate. The hydride
fragment is then transferred to the terminal position (of an
R-olefin) by reductive elimination to yield the isomerized olefin.
Hence, a formal 1,3-H shift is obserVed when the π-allyl
mechanism is actiVe (Figure 4).
On the other hand, the metal hydride mechanism involves a
distinct metal hydride complex already present in the reaction
and is therefore intermolecular: i.e., hydrogen atoms from one
substrate can be transferred to the catalyst onto another substrate
molecule. The olefin inserts into the metal-hydride bond to
form either a primary metal alkyl (thermodynamically favored)
or a secondary metal alkyl that will lead to the isomerized olefin
by â-Η elimination, generating another metal hydride. The
occurrence of both Markovnikov and anti-Markovnikov addi-
1
1
in previous isomerization studies. The composition of the
reaction mixture was identified using 2D NMR and NOE
2
5
experiments. The resulting sample consists of the four
compounds shown in Figure 6: cis- and trans-propenyl methyl
ether (10cis and 10trans) are the isomerization products of the
starting material and represent 70% of the mixture; the remaining
30%, cis- and trans-1,4-dimethoxybut-1-ene (11cis and 11trans),
are the result of isomerization on the metathesis product. Each
of these compounds is a mixture of the four isotopomers derived
from partial deuteration in positions 2 and 3 of the alkylene
ether. Interestingly, position 1 is completely deuterated, while
position 4 does not contain any deuterium. The relative ratios
of each compound and the molar fractions of deuteration in each
2
6
position are given in Figure 6.
The facts that only four products are present, that position 1
is fully deuterated, and that position 4 is not deuterated at all
provide good evidence against deuterium exchange with the
(
24) Werner, H.; Grunwald, C.; Stuer, W.; Wolf, J. Organometallics
003, 22, 1558-1560.
25) The details of NMR and NOE experiments are available in the
Supporting Information.
2
(
(
23) Courchay, F. C. Ph.D. Dissertation, University of Florida, 2005.

Ruthenium-Catalyzed Olefin Metathesis
Organometallics, Vol. 25, No. 26, 2006 6077
Figure 5. Deuterium shifts with a metal hydride addition-elimination mechanism.
Figure 7. Incorporation of hydrogen at C-1 by reversible isomer-
ization.
formed, and its formation is expected to be thermodynamically
irreversible, as it is a vinyl ether. In any case, if the reaction
were reversible, an isomerized olefin, methoxy-1,3-d2-propene
for instance, should be able to insert into a metal-hydride bond
to form a secondary metal alkyl, which could subsequently
undergo a â-D elimination to produce methoxy-3-d-propene
2
Figure 6. Products of the reaction of allyl-1,1-d methyl ether with
complex 2. The molar fractions of each compound are given in
parentheses; the molar fractions of deuteration are indicated at each
position.
(
Figure 7). This transfer of the methylene deuterium to the metal
would automatically bring the percent deuteration at C-1 under
00%, which is not the case. A similar behavior was witnessed
by Krompiec et al. with the hydride complex RuClH(CO)-
1
solvent: i.e., all the deuterium present on the products actually
comes from the starting material, which confirms the validity
of this model study. Therefore, deuterium incorporation at the
C-2 position clearly indicates that the isomerization mechanism
operating here is a metal hydride addition-elimination. In fact,
the relative ratios of deuterium incorporation at positions 2 and
28
(PPh3)3. Under ADMET conditions, the reaction of the allylic
ether N,N-bis[3-(allyloxy)-2-hydroxypropyl]aniline with 2 re-
sulted in complete isomerization, leading to the vinylic ether.1
The distribution of isomerization vs metathesis product is a
result of several factors. Although metathesis is often believed
to proceed more quickly and is therefore the primary reaction,
our conditions are set up in a closed system, which limits its
progression. Once the system is saturated with ethylene, it
reaches equilibrium and metathesis chemistry virtually stops.
Meanwhile, the isomerization reaction continues to consume
starting material and produce a vinyl ether, which can then trap
the metathesis catalyst into a stable, low-reactive Fisher carbene
whose biphosphine analogue is known to thermally decompose
0e
3
reflect the relative rates of Markovnikov (formation of a
secondary metal alkyl) versus anti-Markovnikov addition (for-
mation of a primary metal alkyl) of the metal hydride across
the olefin bond (Figure 6). This ratio is estimated to be about
1
:2 for both 10cis and 10trans, which is consistent with the
preferred formation of the thermodynamically favored primary
metal alkyl. For example, Cramer et al. reported that in the
isomerization of 1-butene by rhodium(III) hydride, addition of
1
6
D to C-2 was 15 times faster than addition to C-3.27 The lower
into the hydride complex 5. ObserVation of a typical color
change from pink to bright yellow, indicatiVe of alkoxy carbene
decomposition, reinforces the argument for the proposed
mechanism. Thus, not only does the metathesis rate decrease
over time but also the isomerization rate increases as the hydride
concentration increases (if we consider that the source of hydride
is related to catalyst decomposition). Hence, only 30% of
metathesis product is formed and is subsequently isomerized
to 10cis and 10trans.
ratio obtained with our allyl ether may be a result of the directing
effect of the oxygen moiety, as described previously by McGrath
and Grubbs during the isomerization of allylic alcohols and
II
13c
ethers by Ru (H2O)6(tos)2.
In the case of 11cis and 11trans, the relative ratios of
deuterium incorporation at C-2 and C-3 are approximately 1:1,
which makes sense, since isomerization occurs on an internal
double bond. Indeed, the metal hydride can only add across
the olefin bond of the metathesis product and forms exclusively
a secondary metal alkyl.
Other Mechanistic Considerations. The olefin isomerization
mechanism operating here cannot involve a â-H shift from an
allylic proton to the metal, as proposed for the reaction of allyl
The complete deuteration of position 1 suggests the irrevers-
ibility of the isomerization reaction. The kinetic product is
2
4
alcohols in the presence of 1. Indeed, the mechanism shown
in Figure 8 involves the stoichiometric consumption of the
isomerization-metathesis catalyst and in our case would have
yielded only up to 0.2 mol % of isomerized starting material.
(
26) The sum of deuterium for each molecule of product 9 and 10 does
not equal 2.00 (number of deuteriums present in one molecule of allyl-
,1-d2 methyl ether) because some deuterium atoms were incorporated in
the ethylene released during the metathesis of allyl-1,1,3-d3 methyl ether.
27) Cramer, R. J. Am. Chem. Soc. 1966, 88, 2272-2282.
1
(28) Krompiec, S.; Antoszczyszyn, M.; Urbala, M.; Bieg, T. Pol. J. Chem.
2000, 74, 737-741.
(

6078 Organometallics, Vol. 25, No. 26, 2006
Courchay et al.
to the case for their phosphine analogues.8,11 Ligand variation
experiments on both the phosphine and carbene ligands have
further reinforced the idea that NHC ligands play a major role
in the isomerization behavior of ruthenium carbene catalysts.23
In light of the results discussed above, we decided to investigate
the eventual participation of NHC groups in the production of
a metal hydride. Speculations evoke a possible ruthenium
insertion into the C-H bond of one of the mesityl groups. To
address this issue, we built a complex that would be comparable
to catalyst 2 in structure and reactivity and would simultaneously
allow us to track its evolution through isotopic labeling
experiments. Complex 8 (Scheme 1) was designed with fully
deuterated o-methyl groups on the NHC aryl rings so that any
sign of ruthenium insertion would be easily detected by NMR
spectroscopy.
Figure 8. â-H abstraction-elimination mechanism proposed for
the reaction of allyl alcohol with 1.
We started with bromo-m-xylene (A) and used an oxidation-
reduction strategy to obtain the deuterated xylene D. Direct
reduction of the diacid B by lithium aluminum deuteride only
yielded unidentified decomposition products. The reaction was
rendered successful by moving to the corresponding diester but
still necessitated controlled conditions to reach reasonable yields
(85%). The alcohol groups of C were substituted to bromine
groups to allow further reduction to the fully deuterated
bromoxylene D. Deuterated 2,6-dimethylaniline was obtained
by substitution of the bromine group through a Grignard reaction
with (trimethylsilyl)methyl azide (TMSDA). The corresponding
ligand salt was then easily prepared according to the regular
procedure. However, we used the chloroform adduct for the
synthesis of the final complex 8, to ensure a better yield and
Figure 9. Total concentration of 9 versus time. Conditions: 2.3
M solution of 9 in C D with 0.2 mol % of 2 at 35 °C for 8 h.
6 6
Nevertheless, this step may be responsible for the formation of
the hydride species involved in the metal hydride isomerization
reaction described above, by reaction of the ruthenium carbene
1
7d
avoid salt contamination.
with vinyl ether, and subsequent decomposition of the Fisher
Both the H and 31P NMR spectra of complex 8 are very
1
carbene hence formed.16
similar to those of catalyst 2, showing the same carbene shift
Kinetic Studies. The relative rates of isomerization versus
metathesis could be measured for allyl-1,1-d2 methyl ether when
the system had reached metathesis equilibrium. Indeed, when
3a
at 19.63 ppm and the same phosphine shift at 29.5 ppm. No
2
hydride trace was detected at this point. The H NMR shows
1
two singlets at 2.35 and 2.75 ppm, representing the aryl rings
in the front (above the benzylidene) and in the back (above the
open coordination site) of the catalyst. Rotation of the NHC
ring about the Ru-C bond is typically slow on the NMR time
scale, as observed with the parent mesityl complex 2, which
exhibits four methyl resonances (in a 3:3:6:6 ratio) at low
temperature. At room temperature the fluxional processes of
the benzylidene induce a broadening of the original signals into
two singlets, as observed here. The crystal structure exhibits a
typical square-pyramidal geometry with the carbene moiety in
the reaction in a closed system was monitored by H NMR, the
starting material quickly disappeared to form 1,4-dimethoxy-
but-2-ene,29 but after 60 min, the rate of formation of metathesis
product became slower until its concentration leveled, indicating
metathesis equilibrium at only 7% conversion. From this point
on, we clearly observed the steady formation of isomerized
starting material. Considering the rate of metathesis to be
negligible and the isomerization of 9 to be irreversible, as shown
above, the isomerization reaction then followed first-order
-
6
kinetics. The rate was calculated at approximately 5.59 × 10
(Figure 9). By considering the concentrations of metathesis
product and starting material at equilibrium, the metathesis
3
1
-1
the apical position (Figure 10). However, almost the entire
molecule is affected with an inherent disorder related to the
existence of two mirror images within the asymmetric unit.
Representative bond lengths and angles of complexes 2 and 8
s
-
6 30
equilibrium constant could also be estimated at 4.04 × 10 ,
a small value that emphasizes the importance of ethylene
removal.
6b
are reported in Table 1 for comparison. Apart from the chlorine
atoms, whose bond lengths to ruthenium vary by about 0.1 Å,
both solid-state structures appear to be very similar. The
deviation of the chlorines is just a result of the global molecular
disorder. In fact, the Ru-Cl bond length averages 2.3929 Å
Now that we know olefin isomerization is promoted by a
metal hydride complex, the question remains: what is the
structure of this hydride species, and what causes its formation?
Realizing the complexity of the question, we first endeavored
to understand the role of NHC ligands in the isomerization
behavior of ruthenium carbene catalysts.
Design, Synthesis, and Characterization. Recent reports
have shown that metathesis catalysts bearing NHC ligands
undergo olefin isomerization to a considerable extent compared
3
2
from the two disordered structures. Since complexes 2 and 8
are so related structurally, it is reasonable to expect similar
catalytic behavior.
Thermal Stability. When a solution of complex 8 in benzene-
d6 was left at 70 °C in a sealed ampule, no sign of ruthenium
2
deuteride was detected by H NMR, even after 21 days, when
(29) Traces of deuterium incorporation at the olefinic bond of the starting
material were also observed, suggesting the presence of a metal hydride at
this early stage of the reaction.
(31) Crystals suitable for X-ray structure determination were obtained
from slow diffusion of pentane into a saturated solution of 8 in benzene at
room temperature.
2
(
30) Metathesis equilibrium constant KM ) [MP]e[C2H4]e/[SM]e
)
2
2
[
MP]e /[SM]e , where MP is the metathesis product, 1,4-dimethoxybut-2-
ene, and SM is the starting material, allyl-1,1-d2 methyl ether.
(32) Crystal data for compound 8 are available as Supporting Information.

Ruthenium-Catalyzed Olefin Metathesis
Organometallics, Vol. 25, No. 26, 2006 6079
Scheme 1. Synthesis of Complex 8a
a
Reagents and conditions: (i) KMnO
4
2 3 3 3
, 50% tBuOH in water, reflux, 16 h (yield 99%); (ii) K CO , CH I, cat. CsCO , DMF, reflux 12 h (yield
8
1%); (iii) LAD, THF, 0 °C, 20 min (yield 85%); (iv) CBr
ether, reflux, 2 h, (b) TMSDA, room temperature, 3 h (yield 55%); (vii) glyoxal, nPrOH, water, room temperature to 60 °C, 6 h (yield 80%); (viii)
NaBH , 1:2 MeOH-THF, room temperature, 16 h, 1 M HCl (yield 85%); (ix) CH(OEt) , reflux, 4 h (yield 60%); (x) NaH, CHCl , room temperature
h (yield 82%); (xi) 1, toluene, 60 °C, 90 min (yield 75%).
4 3 2 2
, PPh , CH Cl (yield 70%); (v) LAD, THF, room temperature (yield 90%); (vi) (a) Mg,
4
3
3
2
Table 1. Selected Bond Lengths (Å) and Angles (deg) for
Complexes 2 and 8
2
8
Bond Lengths
2.085(2)
Ru-C(1)
Ru-C(20)
Ru-Cl(1)
Ru-Cl(2)
Ru-P
2.076(3)
1.844(3)
2.3644(9)
2.4302(10)
2.4298(8)
1.835(2)
2.3988(5)
2.3912(5)
2.4245(5)
Bond Angles
100.24(8)
94.55(5)
C(20)-Ru-C(1)
C(1)-Ru-Cl(1)
C(1)-Ru-Cl(2)
C(20)-Ru-Cl(1)
C(20)-Ru-Cl(2)
Cl(1)-Ru-Cl(2)
C(1)-Ru-P
98.37(11)
88.90(8)
89.65(8)
105.17(9)
87.50(9)
167.32(4)
163.29(7)
97.33(9)
83.26(5)
89.14(7)
103.15(7)
167.71(2)
163.73(6)
95.89(6)
C(20)-Ru-P
Figure 10. X-ray structure of complex 8.
2
3
3
Even though the H NMR does not indicate the presence of
a ruthenium deuteride, a number of deuterated species have
formed. Most signals overlap between 1 and 3 ppm, collapsing
into a broad base, which accounts for the deuterated methyls
of the NHC ligand of the different decomposed ruthenium
the decomposition process was already at an advanced stage.
In fact, observation of the 3 P NMR spectrum indicates that
1
only 4% of the original complex remained. Other typical peaks
have emerged at 10.4, 34.6, 47.0, and 47.4 ppm, along with
what appears to constitute 70% of the phosphorus products
centered at around 72 ppm. The first signal can be attributed to
free PCy3. The peak at 34.6 ppm seems consistent with the
1
3
complexes. The C NMR did not reflect the presence of any
carbonyl compound or carbide. Since no compound could be
isolated, we tested the isomerization behavior of this decom-
position mixture on 1-octene using typical reaction conditions.1
No decomposition product of complex 8 was able to isomerize
olefins to the extent observed with the original metathesis
catalyst (5% of isomers were obtained here vs 76% obtained
+
-
phosphonium salt [PCy3CH2Ph] [Cl] , observed by Grubbs as
1b,34
a byproduct of methylidene 2′ decomposition.20 Interestingly,
we initially attributed the peaks at 47.0 and 47.4 ppm,
representing 15% of the phosphorus products, to the carbonyl
hydride complexes 4 and 5, respectively. These compounds
usually form by reaction of 2 with primary alcohols, but
contamination during the methanol wash following the synthesis
is unlikely to account for the 15% of hydride produced here.
1
1a
1
7b
with catalyst 2 in similar conditions).
Because the methylidene is a key intermediate in most
metathesis reactions, we also examined its decomposition in
situ by reacting 8 with excess ethylene in benzene at 60 °C and
1
Indeed, H NMR only confirmed the presence of 4 and 5
characteristic triplet at -24.25 ppm and a doublet at -24.95
3
5
stirring the mixture for 3 days. Again, no sign of ruthenium
(
ppm) as traces. Therefore, unknown phosphorus compounds
must appear around the same shifts as the carbonyl compounds.
As a matter of fact, no carbonyl stretch was observable by IR
(
34) In a previous report,11a we observed that, after 24 h at 55 °C, a 2.3
M solution of catalyst 2 in neat 1-octene afforded 76% isomers. In the
present case, the catalytic mixture was dried under vacuum once fully
decomposed (no more carbene signal on the H NMR), and 0.5 g of 1-octene
was added. The reaction mixture was allowed to stand at 55 °C for 24 h,
and the isomer content was determined by H NMR by integration of the
internal olefin peak at δ 5.40 ppm.
(
-
1
analysis (usually in the range 1890-1920 cm ). A third
1
1
unidentified hydride is visible on the H NMR as a doublet at
1
2
δ -7.43 ppm ( J ) 50 Hz) that had also been observed by
reaction of 2 with primary alcohols.
35) When the reaction mixture was sampled for 1H NMR, these
conditions led to the complete transformation of benzylidene 8 into the
methylidene, which decomposed rapidly, as indicated by the disappearance
of the methylidene signal at 18.40 ppm.
(
33) The decomposition is indicated by the steady disappearance of the
carbene signal at δ 19.63 ppm.

6080 Organometallics, Vol. 25, No. 26, 2006
Courchay et al.
1
2
Figure 11. (a) H NMR and (b) H NMR spectrum of neat 1-octene with 0.5 mol % of complex 8, after 4 h at 35 °C and 20 h at 55 °C.
2
deuteride was detectable by H NMR, and the same peaks,
observed during the decomposition of the benzylidene, appeared
between 1 and 3 ppm. Interestingly, a singlet at δ 7.14 ppm
Catalytic Behavior under Metathesis Conditions. We
reasoned that if there were formation of a ruthenium deuteride
by metal insertion into one of the methyl bonds, inevitably there
would be deuterium incorporation on the olefin, which should
be easily detectable by NMR.
3
6
indicated the presence of partially deuterated benzene, sug-
gesting the existence of an H/D exchange process between
ligand and solvent, since the only original source of deuterium
in this case was the methyl groups of the carbene ligand
methylidene of complex 8. A similar exchange with solvent was
To avoid any solvent contamination and allow the detection
of both catalyst and substrate on the NMR scale, the experiments
were run in neat substrate, the only deuterium source being
37
observed for the ruthenium dihydride (IMes)Ru(H)2(H2)2PCy3,
complex 8. First, we used 1-octene as our substrate, since we
and other dihydrides have been shown to undergo exchange
processes.38 The species responsible for this H/D exchange may
also be a ruthenium dihydride, which may not be detectable
because of NMR detection/concentration limits. Indeed, many
hydride catalysts are too reactive to be observable spectroscopi-
cally under reaction conditions.
The catalytic mixture, once fully decomposed, was again
tested for isomerization activity using 1-octene under the same
conditions as used previously. Even though the isomerization
extent was increased to 25% isomers, the decomposition
products of complex 8’s methylidene alone did not match the
isomerization behavior observed with catalyst 2 under metathesis
conditions, i.e., in the presence of an olefin, which suggests
that the isomerization process involves more than simply
decomposition products. Other intermediates of the metathesis
cycle must play a role in the formation of the hydride responsible
for this side reaction. For example, Forman et al. have described
the decomposition of a metallacyclobutane through â-H abstrac-
tion to form a ruthenium hydride and explained the appearance
of different alkenes from the metathesis of ethylene with [Ru]-
11
have based previous isomerization studies on this model.
A
0
.5 mol % solution of complex 8 in 1-octene was monitored
1
2
by H and H NMR at 35 °C for 4 h. These reaction conditions
will ensure significant amounts of both metathesis and isomer-
11
1
2
ization. The H and H spectra were systematically superim-
posed to match chemical shifts, since no internal standard was
available. Because of the chemical simplicity of the substrate,
the distinction between isomers and metathesis products is
1
virtually impossible by standard H NMR. Still, we know that
under the present conditions metathesis occurs much more
quickly than the isomerization process.
Shortly after the experiment was started, an internal olefin
signal appeared on the proton spectrum, marking the beginning
of the metathesis process, while the deuterium spectrum only
showed the two singlets from the catalyst deuterated methyl
groups. However, after 4 h, several other peaks became visible
2
in the H spectrum, confirming incorporation of deuterium at
various positions of the olefin backbone. Figure 11 shows the
1
2
H and H NMR spectra taken after 24 h of reaction. A multiplet
corresponding to the terminal olefin suggests the formation of
CHDdCH(CH2)5CH3. Two other vinyl signals at δ 5.75 and
5.38 ppm, though smaller, evidence deuterium incorporation at
the C-2 position of 1-octene and on an internal double bond,
probably from an isomer of either 1-octene or 7-tetradecene.
These observations are consistent with the presence of a metal
deuteride, which is further supported by the emergence of peaks
in the allylic region (around δ 2.0 ppm) followed by peaks in
the alkane region (at δ 1.30 and 0.90 ppm), supposing that the
isomerization reaction progresses along the olefin backbone to
produce 1-d-2-, 1-d-3-, and 1-d-4-octene and 8-d-3-, 8-d-2-, and
3
9
1
′.
Hence, we decided to further investigate hydride formation
in the presence of olefinic substrates to reflect true metathesis
conditions.
(
36) The deuterium content observed here is much superior to the levels
naturally found in benzene.
37) Giunta, D.; H o¨ lsher, M.; Lehmann, C. W.; Mynott, R.; Wirtz, C.;
Leitner, W. AdV. Synth. Catal. 2003, 345, 1139-1145.
38) (a) Burtling, S.; Mahon, M. F.; Paine, B. M.; Whittlesey, M. K.;
Williams, J. M. J. Organometallics 2004, 23, 4537-4539. (b) Abdur-Rashid,
K.; Fedorkiw, T.; Lough, A. J.; Morris, R. H. Organometallics 2004, 23,
6-94.
39) van Rensburg, W. J.; Steynberg, P. J.; Meyer, W. H.; Kirk, M. M.;
Forman, G. S. J. Am. Chem. Soc. 2004, 126, 14332-14333.
(
(
8
-d-1-octene, for example (Figure 12). The formation of the
8
metal alkyl may be fast relative to the decomplexation of the
olefin; thus, the same molecule of substrate can undergo several
(

Ruthenium-Catalyzed Olefin Metathesis
Organometallics, Vol. 25, No. 26, 2006 6081
Figure 12. Formation of 1-octene isomers catalyzed by a ruthenium deuteride complex.
concentration is simply too low. However, we reasoned that if
there was indeed formation of a ruthenium deuteride complex
in sufficient amounts, then for each deuterium atom incorporated
onto a substrate molecule, the corresponding ruthenium hydride
species would be produced, according to the metal hydride
addition-elimination mechanism. Therefore, this hydride com-
1
Figure 13. Complex 9.
plex may be visible by H NMR as the isomerization process,
producing the D/H exchange, goes on. A doublet appeared at δ
transformations before being released.13f Despite the large excess
of 1-octene, it is also possible that some of the deuterated
products come from isomerization of the metathesis product,
although this cannot be quantified. It is worth noting that the
alkyl deuteron signals are usually broader and flatter than those
of their vinyl analogues because of multiple couplings. For this
reason, the methyl signal expected for 1-d-2-octene at δ 1.60
2
-
20.36 ppm with JHP ) 17.8 Hz, characteristic of a hydride
situated trans to an empty coordination site and cis to a
17d
phosphine. The same signal had been observed during the
decomposition of the methylidene, and interestingly, only the
methylidene remains at this point, implying that all the catalyst
has either been activated by the olefin or decomposed. However,
the information collected thus far does not allow us to conclude
whether this hydride complex is in fact responsible for olefin
isomerization under metathesis conditions.
The same experiment was performed with allyl methyl ether,
which was used to evaluate the isomerization mechanism. We
thought that dealing with an irreversible isomerization reaction
would facilitate product identification, being limited by NMR
spectroscopy in our interpretation. Again, 0.5 mol % of catalyst
was used to provide sufficient deuterium content for NMR
detection.
2
ppm only becomes visible by H NMR after longer reaction
times. Nevertheless, deuterium atoms appear to preferentially
incorporate at the C-1 position, suggesting a predominant
Markovnikov addition of the metal deuteride across the double
bond. When the temperature was raised to 55 °C, the concentra-
tion of these deuterated terminal olefins increased even more
quickly. According to fundamental entropy thermodynamics,
decomplexation of the olefin must occur to a greater extent at
higher temperatures, therefore generating more olefins at a lower
degree of isomerization. Indeed, after 20 h at 55 °C, the signal
for the allylic methyl of 1-d-2-octene at 1.60 ppm has increased
significantly and is now larger than other allylic or alkyl signals
After 4 h at 35 °C, about 30% of metathesis product had
1
formed, with no isomerization product detectable in the H
NMR, the high concentration of ether favoring the bimolecular
metathesis process over the unimolecular isomerization reaction.
(Figure 11).
Aside from the signals of the deuterated methyl groups of
2
The H NMR spectrum shows deuterium incorporation mostly
the metathesis catalyst at δ 2.40 ppm, two broad singlets arise
at 2.25 and 2.65 ppm. This indicates a transformation of the
original complex. However, none of the complexes identified
so far from catalyst decomposition could match our observations.
Even though the multiple peaks between δ 2 and 3 ppm recall
hydride complex 6 formed by thermal decomposition of the
methylidene 2′, they do not account for the formation of the
4
0
at the terminal carbon of the olefin, as allyl-3-d methyl ether,
1
the signal matching perfectly the pattern of the H spectrum
Figure 14). Deuterated ethylene is also present in significant
(
amounts, more likely by metathesis of allyl-3-d methyl ether.
Again, this is consistent with the presence of a metal deuteride
complex. Smaller resonances at 1.68 ppm indicate the formation
of isomerized product, as 3-d-1-methoxypropene.
20
metal deuteride evidenced in this experiment. Likewise, we
do not observe the characteristic signal from the Ru-CD2Ar
moiety that would have probed the presence of 1 equiv of
complex 7. Of course, this does not exclude the possibility of
a different C-D activation. In fact, hydride complex 9 shown
in Figure 13, formed by insertion of ruthenium into the C-H
bond of a mesityl group in the presence of alkenes, exhibits a
broad “triplet” at 2.76 ppm from the protons of the Ru-CH2Ar
moiety. Other examples of C-H activation from ruthenium
carbenes could partially match these signals; however, these
assignments must remain speculations as long as the components
of the mixture are not isolated.
Surprisingly, a triplet at 3.13 ppm and a doublet at 3.72 ppm
suggest the presence of deuterons at the C-4 and C-1 positions
of the starting material (less than 0.3%), respectively. A classical
metal hydride mechanism cannot explain deuteration at these
positions, nor can a π-allyl mechanism. Hence, an alternate
mechanism must be proceeding competitively; however, to a
much lower extent. The H/D exchange process observed during
the decomposition of complex 8’s methylidene may also be
operating here. If this is the case, the introduction of a
nonisomerizable olefin, such as 3,3-dimethylbutene, should
result in the deuteration of every position and, more particularly,
of the methyl groups.
As a matter of fact, no ruthenium deuteride species was
detected, even after longer reaction times. The complex
responsible for the isomerization process probably reacts more
quickly than the NMR time scale allows for detection, or its
(40) Due to the large excess of allyl methyl ether, we suppose that at
early reaction times the molecules are only monodeuterated, since they are
not likely to react again with another ruthenium deuteride.

6082 Organometallics, Vol. 25, No. 26, 2006
Courchay et al.
2
3
37
sp rather than sp C-H bonds. Even though the event of the
formation of a dihydride complex from a ruthenium carbene in
1
5
the absence of dihydrogen may seem unlikely, the reactivity
of the phosphine-free methylidene offers many possibilities. For
instance, Stradiotto et al. reported a reversible R-H elimination
allowing the interconversion of RudC and Ru-alkyl species.43
The electrophilicity of the methylidene carbon also renders it
highly sensitive to nucleophilic attack. In fact, Grubbs et al.
proposed that the first step toward formation of bimetallic
complex 6 involves nucleophilic attack by free phosphine to
form a ruthenium alkyl complex. Certainly, the hypothesis of
a ruthenium dihydride is not ruled out by the spectroscopy
analysis of 2′ decomposition. One of the observed hydride
signals may very well be one of a dihydride. Finally, the same
dihydride complex could be responsible for both the exchange
process and the isomerization reaction.
4
4
2
0
Figure 14. 2H NMR spectrum of neat allyl methyl ether with 0.5
mol % of complex 8, after 4 h at 35 °C.
The rate of the H/D exchange process apparently depends
on the substrate. Indeed, we observe greater deuterium incor-
poration on the methyls of 3,3-dimethylbutene than on allyl
methyl ether. The extensive production of the methylidene
complex, inherent to the metathesis of olefins containing allylic
methyls, accelerates the rate of catalyst decomposition, which
is probably in direct correlation with the formation of the species
promoting the exchange process.
Conclusions
A deuterium labeling study was undertaken to determine the
mechanism of olefin isomerization during the metathesis reac-
tions catalyzed by a second-generation Grubbs catalyst. The
reaction of allyl-1,1-d2 methyl ether with 2 at 35 °C was
Figure 15. 2H NMR spectrum of neat 3,3-dimethylbutene with
0.5 mol % of complex 8, after 4 h at 35 °C.
1
2
followed by H and H NMR spectroscopy. The evidence of
deuterium incorporation at the C-2 position allowed us to
conclude that a metal hydride addition-elimination mechanism
was operating under these conditions. When the reaction was
conducted in a closed system, the isomerization rate was
To look into this potential H/D exchange process, we repeated
the experiment with 3,3-dimethylbutene for its inherent inability
to isomerize. We reasoned that if there were indeed formation
of a “classical” ruthenium deuteride, we would observe deute-
rium incorporation exclusiVely at the C-1 and C-2 positions. In
addition, the metathesis reaction of olefins containing allylic
methyls is known to proceed at a slow rate, if at all. In fact, the
cross-metathesis of 3,3-dimethylbutene with 1 only resulted in
an accumulation of the methylidene complex by nonproductive
metathesis.41 Accordingly, after 4 h at 35 °C, metathesis
conversions merely reached 10%.
-6 -1
calculated at approximately 5.59 × 10 s , while the equi-
-6
librium constant was about 4.04 × 10 .
The next step was an attempt to identify the eventual role of
the NHC ligand in the formation of the hydride species
responsible for the isomerization reaction. We synthesized
complex 8, an analogue of 2, bearing deuterated o-methyl groups
on the aromatic rings of the NHC ligand. Its decomposition
afforded several metal hydride species; however, no deuteride
2
2
Figure 15 illustrates the corresponding H NMR. Even though
complex was detected by H NMR spectroscopy. The meth-
there are signs of deuteration on the double bond (signals at δ
ylidene analogue 8′ was also subjected to thermal decomposition
but again did not afford any detectable ruthenium deuteride
complex. The isomerization activities of the catalytic decom-
position mixtures were tested with benchmark 1-octene but did
not match the isomerization rates observed with 2 under similar
metathesis conditions, proving that thermal decomposition
products from this catalyst do not promote the isomerization
reaction. Hence, other metathesis intermediates, such as the
metallacyclobutane, must be involved in the isomerization
process. Indeed, when complex 8 was reacted with 1-octene, a
variety of deuterated olefins were produced, indicating the
presence of a deuteride complex, albeit not observable spec-
troscopically. Finally, reaction of 8 with allyl methyl ether and
3,3-dimethylbutene actually revealed the existence of a competi-
tive H/D exchange process between the CD3 groups on the NHC
ligand and the C-H bonds of the substrate. We propose that
5.70 and 4.70 ppm), most displaced deuterons are on the methyl
groups, as indicated by the major peak at 0.86 ppm. This is
clearly the result of an H/D exchange process, although not
necessarily direct, between the methyl groups of the NHC ligand
and the C-H bond of the substrate. It is likely that the deuterium
atoms are transferred to the metal center from the NHC methyl
groups and then onto the substrate. Again, this type of transfer
could very well be the result of ruthenium insertion into the
methyl C-D bond, the cyclometalation of a substituent on a
nitrogen-ligated carbene on ruthenium being a fairly common
reaction, albeit not well understood.17d,21,38,42 Interestingly, the
ruthenium dihydride (IMes)Ru(H)2(H2)2PCy3 promotes a similar
exchange from the mesityl methyl groups to a variety of
aromatic compounds, although preferentially exchanging with
(
(
41) Ulman, M.; Grubbs, R. H. Organometallics 1998, 17, 2484-2489.
42) (a) Owen, M. A.; Pye, P. L.; Piggott, B.;Capparelli, M. V. J.
(43) Rankin, M. A.; McDonald, R.; Ferguson, M. J.; Stradiotto, M.
Angew. Chem., Int. Ed. 2005, 44, 3603-3606.
(44) Hansen, S. M.; Rominger, F.; Metz, M.; Hofmann, P. Chem. Eur.
J. 1999, 5, 557-566.
Organomet. Chem. 1992, 434, 351-362. (b) Hitchcock, P. B.; Lappert, M.
F.; Pye, P. L.; Thomas, S. J. Chem. Soc., Dalton Trans. 1979, 88, 1929-
1
942.

Ruthenium-Catalyzed Olefin Metathesis
Organometallics, Vol. 25, No. 26, 2006 6083
Figure 16. 1H and 13C chemical shift assignments in the com-
pounds 10cis, 10trans, 11cis, and 11trans.
Figure 17. Molar fractions of compounds 10cis, 10trans, 11cis,
and 11trans in the product (under the compound number) and molar
fractions of deuteration at each position. The number on top was
based on the proton spectrum and the second number on the
deuterium spectrum. The number on the bottom for the molar
fraction of deuteration was based on the integral of the proton signal
at that position and the composition determined in the deuterium
spectrum, and it is the most reliable.
the H/D exchange is promoted by a ruthenium dihydride
intermediate whose formation is closely related to the meth-
ylidene decomposition. Further studies must be conducted to
isolate and characterize these hydride complexes. Meanwhile,
hydride traps should be used during metathesis reactions
catalyzed by ruthenium carbenes in order to prevent competitive
isomerization.45
1
1cis and 11trans. All of these signals display an extra coupling
Experimental Section
with deuterium: triplet in a 1:1:1 ratio.
General Considerations. Routine 1H NMR (300 MHz) and 13C
NMR (75 MHz) measurements were recorded on either a Mercury
Series or Varian VXR-300 NMR superconducting spectrometer.
Chemical shifts are reported in ppm relative to tetramethylsilane
These signals present in the gHMBC spectrum (Figure 19) a
one-bond coupling to a carbon in the range 96-102 ppm and two
long-range couplings, one with a carbon at 9-28 ppm and one
with a carbon at 147-149 ppm. This latter carbon displays a triplet
in f1, consistent with a one-bond coupling with deuterium, and also
a long-range coupling to protons at 3.16-3.18 ppm, protons which
are on carbons at 55-60 ppm. The protons on the carbons at 9-28
ppm are in the range 1.43-2.41 ppm and couple with both the
carbons at 96-102 ppm and with the carbons at 147-149 ppm.
These proton-carbon couplings, together with the characteristic
chemical shifts, identify the methyl-1-propenyl ether fragment in
all four compounds. In the compounds 11cis and 11trans, the
aliphatic protons at 2.06 and 2.41 ppm couple with carbons at 73.1
and 72.0 ppm, respectively. Both of these carbons carry no protons
and couple with the protons of a methoxy group. The NOESY
spectrum reveals NOE’s between the alkene proton and the methoxy
protons of the methyl vinyl ether fragment in compounds 10trans
and 11trans only; therefore, in these compounds the double bond
is E. In these compounds the proton-deuterium coupling across
the double bond is ca. 1.8 Hz, while in the Z compounds 10cis and
(
TMS) or residual proton from the solvent. All organometallic
spectra were recorded in d -benzene, previously dried over Na/K
6
alloy and degassed by three freeze-pump-thaw cycles.
All materials were purchased from Aldrich and used as received,
except for the THF and ether that were previously dried over a
4
6
47
catalyst bed. Complex 2, allyl-1,1-d
2
methyl ether, and (tri-
methylsilyl)methyl azide (TMSMA)48 were synthesized according
to the literature. Allyl-1,1-d methyl ether was distilled, degassed
by three freeze-pump-thaw cycles, and stored in a glovebox.
Deuterium Labeling Study with Allyl-1,1-d Methyl Ether.
In the drybox, 4 mg (0.005 mmol) of complex 2 was dissolved
with 0.8 mL of C or toluene-d in a J. Young valve NMR tube,
and this solution was then layered with 185 mg (2.5 mmol) of allyl-
,1-d methyl ether. The tube was shaken right before being loaded
2
2
D
6 6
8
1
2
into the NMR instrument.
NMR spectra were recorded on a Varian Inova spectrometer
equipped with a 5 mm indirect detection probe, operating at 500
11cis, it is ca. 0.9 Hz.
1
13
The carbon spectrum (Figure 20) confirms the lack of deuterium
MHz for H and at 125 MHz for C. The temperature was 35 °C.
Chemical shifts are reported in ppm relative to TMS. The residual
methyl signal of the solvent was used as reference (2.09 ppm for
on the methoxy groups and complete deuteration at position 1 of
the alkene fragment, for all four compounds. In compounds 11cis
and 11trans, the signals at 73.1 and 72.0 ppm could barely be seen,
as broad multiplets, confirming that these carbons carry only
deuterium. With two positions with partial deuteration, compounds
1
13
H and 20.4 ppm for C).
The sample is a mixture of four compounds, 10cis, 10trans,
1cis, and 11trans, shown in Figure 16. Each of these compounds
1
10cis, 10trans, 11cis, and 11trans are each present as four
is a mixture of the four isotopomers resulting from partial
deuteration in positions 2 and 3 of the alkylene ether. Positions 1
and 4 are completely deuterated. The relative ratios of the four
compounds and the molar fractions of deuteration in each position
are given in Figure 17.
The proton spectrum displays in the region 4.70-4.20 ppm the
signals of a mixture of four compounds (Figure 18): quartets for
the compounds 10cis and 10trans and triplets for the compounds
isotopomers. This is why the carbons in position 3 of the alkene
fragments display a pattern of eight linessthe one at higher field
is the H3,H2 isotopomer, followed by the line of the H3,D2
isotopomer. The next six lines are less intense and belong to the
triplets of the D3,H2 and D3,D2 isotopomers. The carbon chemical
shifts in Figure 16 are those of the H2,H3 isotopomer.
The molar fractions of the four compounds in the mixture and
the fractions of deuterium substitution at each position were
determined by integration of the signals in both the proton and the
deuterium spectra. They are given in Figure 17, as the first and
second sets of values, respectively. The spectra were taken at 70
°C, when the separation of the signals of interest was optimal. In
the proton spectrum, the signals of the methoxy groups were used,
since there is no deuteration of these groups. Four signals could be
integrated separately: 3.22-3.20 (10cis), 3.20-3.18 (10trans +
11cis + 11trans), 3.14-3.12 (11cis), and 3.12-3.11 ppm (11trans).
(45) A recent paper shows the use of benzoquinone to prevent olefin
isomerization during metathesis reactions: Hong, S. H.; Sanders, D. P.;
Lee, C. W.; Grubbs, R. H. J. Am. Chem. Soc. 2005, 127, 17160-17161.
(46) Scholl, M.; Ding, S.; Lee, C.; Grubbs, R. H. Org. Lett. 1999, 1,
9
2
2
53-956.
(
35.
47) McGrath, D. V.; Grubbs, R. H. Organometallics 1994, 13, 224-
(48) Nishiyama, K.; Tanaka, N. J. Chem. Soc., Chem. Commun. 1983,
2, 1322-1323.

6084 Organometallics, Vol. 25, No. 26, 2006
Courchay et al.
2
Figure 18. Expansion of the proton spectrum, the CH region.
Figure 19. Expansion of the gHMBC spectrum.
Figure 20. Expansion of the carbon spectrum, C3 region.
The separation of the last two signals is marginal, and for this reason
the values for the relative ratio of the four compounds determined
in the proton spectrum are considered to be less precise than the
values determined in the deuterium spectrum. Four signals in the
deuterium spectrum were used: 6.30-6.12 (position 1 in 10trans
of toluene. We consider that the best approach is to determine the
relative ratios of compounds in the deuterium spectrum and then
to use these ratios and the integrals from the proton spectrum to
calculate the fraction of deuteration at each position. Values
determined this way are presented in the third row in Figure 17.
+
11trans), 5.78-5.62 (position 1 in 10cis + 11cis), 3.30-3.20
Deuterium Labeling Study with Allyl-1,1-d
In the drybox, 4 mg (0.005 mmol) of complex 2 was dissolved
with 0.8 mL of C in a J. Young valve NMR tube, and this
solution was then layered with 185 mg (2.5 mmol) of allyl-1,1-d
2
Methyl Ether.
(
position 3 in 11cis), and 3.20-3.09 ppm (position 3 in 11trans).
The fraction of deuteration at positions 2 and 3 was determined
6 6
D
in both the proton and the deuterium spectra. The separation of the
signals was poorer in the deuterium spectrum, in which also the
signal of position 3 in 11trans was overlapping the methyl signal
2
methyl ether. The tube was shaken right before being loaded in
1
the NMR instrument. The reaction was monitored by H NMR at

Ruthenium-Catalyzed Olefin Metathesis
5 °C, and the conversions were calculated by integration of the
Organometallics, Vol. 25, No. 26, 2006 6085
3
dropwise at 0 °C. After the addition was complete, the slurry was
peaks at δ 4.98 and 5.72 ppm for the starting material and metathesis
product, respectively, against the solvent residual peak, every 120
s.
stirred for 20 min. The reaction mixture was quenched and worked
up following the standard procedure,50 to yield 20.6 g (0.093 mol,
1
85%) of a white powder (mp 152-154 °C). H NMR (pyridine-d
5
,
2
X-ray Studies. Experimental data: C40
H
61Cl
2
N
2
PRu, M
r
) 820.9,
300 MHz): δ 7.95 (d, 2H, meta CH, JH,H ) 7.2 Hz), 7.50 (t, 1H,
3
13
triclinic, P 1h , a ) 11.706(18) Å, b ) 12.316(19) Å, c ) 16.042(3)
Å, R ) 87.77(3)°, â ) 70.83(2)°, γ ) 70.93(3)°, V ) 2058.5(6)
para CH, JH,H ) 7.5 Hz), 5.14 (s, 2H, OH). C NMR (pyridine-
, 75 MHz): δ 140.9, 126.3, 125.6, 112.1, 62.2 (q, CD OH); MS
(LRMS/EI): calcd for C BrO , m/z 221, found, m/z 221.
d
5
2
3
-3
Å, Z ) 2, Dcalcd ) 1.32 g cm , Mo KR radiation (λ ) 0.710 73
Å), T ) 173 K.
8
H
5
D
4
2
(d) Synthesis of 2-Bromo-R,R′-dibromo-m-xylene-d . 2-Bromo-
4
Data were collected at 173 K on a Siemens SMART PLAT-
FORM equipped with a CCD area detector and a graphite
monochromator utilizing Mo KR radiation (λ ) 0.710 73 Å). Cell
parameters were refined using up to 8192 reflections. A full sphere
of data (1850 frames) was collected using the ω-scan method (0.3°
frame width). The first 50 frames were remeasured at the end of
data collection to monitor instrument and crystal stability (maximum
correction on I was <1%). Absorption corrections by integration
were applied on the basis of measured indexed crystal faces.
The structure was solved by direct methods in SHELXTL6 and
refined using full-matrix least squares. The non-H atoms were
treated anisotropically, whereas the hydrogen atoms were calculated
in ideal positions and were riding on their respective carbon atoms.
The structure has two disorders. The first disorder is a minor one
where the C2-C3 moiety is refined in two parts with their site
occupation factors independently refined. The second disorder is a
major one involving the triphenylphosphine group and the C20
ligand according to a pseudo-mirror symmetry passing through Ru
and C1 and bisecting the C2-C3 bond. A total of 539 parameters
were refined in the final cycle of refinement using 13 498 reflections
with I > 2σ(I) to yield R1 and wR2 values of 3.77% and 8.40%,
1,3-bis(hydroxy-d -methyl)benzene (20.6 g, 0.093 mol) was sus-
2
pended in a solution of CBr (71.2 g, 0.214 mol, 2.3 equiv) in
4
dichloromethane. Triphenylphosphine (57.4 g, 0.219 mol, 2.35
equiv) was added portionwise at room temperature to avoid
overheating. After the addition was complete, the clear solution
was precipitated in ether and filtered. As the solvent was evaporated,
any remaining triphenylphosphine was further filtered. The crude
product was purified via column chromatography using 5:1 hex-
anes-ethyl acetate to yield 22.5 g (0.065 mol, 70%) of a white
1
powder (mp 85-88 °C). H NMR (CDCl , 300 MHz): δ 7.41 (d,
3
2
3
2H, meta CH, J ) 7.3 Hz), 7.28 (t, 1H, para CH, J ) 7.2
H,H
H,H
Hz). 13C NMR (CDCl , 75 MHz): δ 138.4, 128.4, 127.9, 126.9,
3
5
33.9 (q, CD Br, J ) 18.9 Hz); MS (LRMS/EI): C H D Br )
2
C,D
8
3
4
3
347, found 347.
(
e) Synthesis of 2-Bromo-m-xylene-d
6
(D). In a flame-dried
flask, 2-bromo-R,R′-dibromo-m-xylene-d
4
(22.5 g, 0.065 mol) in
THF was added dropwise to a slurry of lithium aluminum deuteride
2.73, 0.065 mol) in dry THF at room temperature. Then, the
(
reaction mixture was quenched and worked up following the
50
standard procedure, to yield 11.2 g (0.058 mol, 90%) of a colorless
1
13
liquid. H NMR (CDCl
CDCl , 75 MHz): δ 138.4, 128.3, 127.8, 126.7, 23.2 (q, CD
C,D ) 18.8 Hz). MS (LRMS/EI): calcd for C Br, m/z 191;
found, m/z 191.
f) Synthesis of 2,6-dimethyl-d
three-neck round-bottom flask equipped with a condenser, 11.2 g
0.058 mol) of 2-bromo-m-xylene-d was added dropwise to a
3
, 300 MHz): δ 7.05 (m, 3 Hz). C NMR
2
49
respectively. Refinement was done using F .
(
J
3
2
Br,
Synthesis of Complex 8. (a) Synthesis of 2-Bromoisophthalic
Acid (B). In a round-bottom flask equipped with a condenser, 25.0
5
8 3 6
H D
g of bromoxylene (0.135 mol) was dissolved in 200 mL of 1:1
(
3
-aniline (E). In a flame-dried
t
4
BuOH-water. Two equivalents of KMnO (42.7 g, 0.270 mol)
was added portionwise, and the solution was refluxed for 4 h. After
the mixture was cooled, 2 equiv more of KMnO was added, and
(
6
4
mixture of 1.71 g (0.070 mol, 1.2 equiv) of magnesium in 50 mL
of anhydrous ether or a minimum amount. The solution was refluxed
upon addition and was subsequently maintained at reflux for 2 h.
After the reaction mixture was cooled to room temperature, a 1.2
M solution of TMSMA in ether was dripped into it, and this mixture
was then stirred at room temperature for an additional 3 h. Water
was added to quench the reaction mixture, and the ethereal phase
the solution was refluxed overnight. The mixture was filtered
through a Celite bed while still hot, and the filtrate was concentrated
to one-third of its original volume. Concentrated H SO was added
2 4
slowly until a white precipitate formed. Filtration and drying under
vacuum afforded 32.7 g (0.133 mol, 99%) of a white powder (mp
1
>
250 °C). H NMR (CD
3
OD, 300 MHz): δ 7.73 (d, 2H, meta
2
3
CH, JH,H ) 7.6 Hz), 7.48 (t, 1H, para CH, JH,H ) 7.8 Hz), 5.32
s, 2H, COOH). 13C NMR (CD
OD, 75 MHz): δ ) 169.9, 137.7,
38.5, 128.4, 118.3. LRMS/EI: calcd for C BrO , m/z 245; found,
m/z 245.
b) Synthesis of Dimethyl 2-Bromoisophtalate. 2-Bromoisoph-
thalic acid (32.7 g, 0.133 mol) was dissolved in DMF. Two
equivalents (37.0 g, 0.270 mol) of K CO and 38.0 g (0.27 mol,
.0 equiv) of methyl iodide were added to the solution, as well as
was extracted, washed with brine, and dried over MgSO
product was purified through column chromatography in CH
to afford 4.09 g (0.032 mol, 55%) of an orange-yellow liquid. H
4
. The crude
Cl
(
1
3
2
2
8
H
5
4
1
2
NMR (CDCl
Hz), 7.19 (t, 1H, para CH, JH,H ) 7.5 Hz), 4.46 (s, 2H, NH
NMR (CDCl , 75 MHz): δ )142.8, 128.4, 121.7, 118.2, 17.4 (q,
, JC,D ) 18.8 Hz); MS (LRMS/EI): calcd for C N, m/z
27; found, m/z 127.
3
g) Synthesis of Glyoxalbis(2,6-dimethyl-d -phenyl)imine. The
3
, 300 MHz): δ ) 7.45 (d, 2H, meta CH, JH,H ) 7.8
(
3
13
2
); C
3
2
3
5
CD
3
8 5 6
H D
2
1
3
a catalytic amount of CeCO . The reaction mixture was stirred
(
overnight at room temperature and quenched with acidic water until
all the base dissolved. The aqueous solution was extracted with
ether. Afterward the organic layer was washed with water, dried
51
procedure was modified from the literature. Under argon, a
mixture of 2.12 g (0.014 mol) of a 40% aqueous solution of glyoxal,
2
4
.3 mL of n-propanol, and 5.8 mL of water was combined with
.09 g (0.032 mol, 2.2 equiv) of 2,6-dimethyl-d -aniline in 20 mL
4
over MgSO , and concentrated to a yellow oil (30.0 g, 0.110 mol,
1
2
6
8
)
3
1%). H NMR (CDCl , 300 MHz): δ 7.68 (d, 2H, meta CH, JH,H
of n-propanol. The solution was stirred at room temperature for 4
h and then at 60 °C for 2 h. Upon addition of 30 mL of water, a
yellow solid precipitated. The mixture was chilled in ice to allow
3
7.4 Hz), 7.40 (t, 1H, para CH, JH,H ) 7.8 Hz), 3.94 (s, 3H,
COOCH
13
3
). C NMR (CDCl
3
, 75 MHz): δ 166.6, 135.2, 132.1,
BrO , m/z 273;
1
27.1, 118.7, 52.6. MS (LRMS/EI): calcd for C10
found, m/z 273.
c) Synthesis of 2-Bromo-1,3-bis(hydroxy-d
C). In a flame-dried round-bottom flask immersed in an ice bath,
.23 g (0.219, 2.0 equiv) of lithium aluminum deuteride was
suspended in 200 mL of dry THF. Dimethyl 2-bromoisophtalate
30.0 g, 0.110 mol) diluted in 20 mL of dry THF was added
H
9
4
further precipitation, and 3.09 g (0.011 mol, 80%) was collected
1
by filtration (mp 145-146 °C). H NMR (CDCl
3
, 300 MHz): δ
(
2
-methyl)benzene
2
8
.13 (s, 2H, CHdN), 7.12 (d, 4H, meta CH, JH,H ) 7.4 Hz), 6.98
(
9
(
50) Fieser, L. F.; Fieser, M. Reagents for Organic Synthesis; Wiley:
New York, 1967; Vol. 1, pp 581-585.
51) Arduengo, A. J., III; Krafczyk, R.; Schmutzler, R.; Craig, H. A.;
(
(
Goerlich, J. R.; Marshall, W. J.; Unverzagt, M. Tetrahedron 1999, 55,
14523-14534.
(49) SHELXTL6; Bruker-AXS, Madison, WI, 2000.

6
086 Organometallics, Vol. 25, No. 26, 2006
Courchay et al.
(
1
t, 2H, para CH, 3JH,H ) 7.7 Hz). 13C NMR (CDCl
3
, 75 MHz): δ
, JC,D ) 18.9 Hz).
(2.35 mmol, 2.2 eq) of ligand G. The mixture was dissolved in 25
mL of toluene and stirred at 60 °C for 90 min under argon. The
solvent was then evaporated, and the brown residue was washed
with methanol (2 × 25 mL) and pentane (50 mL). Finally, the pink
5
63.7, 150.1, 128.5, 126.5, 125.0, 17.4 (q, CD
3
+
MS (HRMS/EI): calcd for C18
found, m/z 299.2282.
H D N
8 12 2
[M + Na] , m/z 299.2272;
(
h) Synthesis of Bis((2,6-dimethyl-d
Dihydrochloride. A suspension of 3.09 g (0.011 mol) of the
glyoxalbis(2,6-dimethyl-d -phenyl)imine in 45 mL of 2:1 THF-
methanol was treated with 1.70 g (0.044 mol, 4.0 equiv) of sodium
borohydride at room temperature. The mixture was stirred for 16
h at room temperature and refluxed for 1 h. Upon cooling, 45 mL
of ice-water was added, followed by 45 mL of 3 M hydrochloric
acid. A white solid precipitated, which was filtered and washed
3
-phenyl)amino)ethane
solid was dried under vacuum to yield 0.693 g (0.83 mmol, 78%)
1
of complex 8. H NMR (toluene-d
8
, 300 MHz): δ 19.57 (s, 1H,
3
RudCH), 7.17 (t, 2H, para CH, JH,H ) 7.2 Hz), 7.13-6.94 (m,
3
3
9H, para, meta CH), 6.54 (t, 1H, JH,H ) 7.3 Hz), 3.43-3.15 (m,
1
3
4H, NCH
NMR (toluene-d
125.8, 125.1, 32.4, 32.2, 30.3, 28.6, 28.5, 27.0, 20.8 (q, CD
) 19.7 Hz). 31P NMR (toluene-d
, 300 MHz): δ 29.69. MS
(HRMS/EI): calcd for C44
found, m/z 831.3648. Anal. Calcd for C44
2
CH
2
N), 2.41 (q, 3H), 1.55 (m, 15H), 1.08 (m, 15H).
C
8
, 75 MHz): δ 147.8, 129.5, 128.9, 128.6, 127.9,
5
3
, JC,D
8
+
with ether and acetone. Yield: 3.35 g (9.9 mmol, 85%); mp >250
°
CH), 3.60 (s, 4H, NCH
H
48
12
D N
2
Cl
2
PRu [M] , m/z 831.3656;
Cl PRu: C,
1
C. H NMR (DMSO-d
6
, 300 MHz): δ 7.12 (s, 6H, para, meta
H
48
D
12
N
2
2
). 1 C NMR (DMSO-d
3
, 75 MHz): δ 135.5,
, JC,D ) 18.8 Hz). MS (LRMS/
64.38; H, 7.49; N, 3.41. Found: C, 64.80; H, 7.39; N, 3.31.
Thermal Decomposition of Complex 8. In the drybox, 32 mg
(0.038 mmol) of complex 8 was dissolved with 0.8 mL of C D in
6 6
a glass ampule, which was then equipped with a Swedgelock
connected to a Teflon Schlenk valve. The system was taken out of
the glovebox, and the ampule was sealed under argon. The ampule
was left in a 70 °C bath for 21 days and shaken regularly. It was
then taken back into the drybox and transferred into a J. Young
valve NMR tube.
2
6
5
1
31.7, 129.9, 127.9, 17.8 (q, CD
EI): calcd for C18 Cl
i) Synthesis of 1,3-Bis(2,6-dimethyl-d
Chloride (F). A mixture of 3.35 g (9.9 mmol) of bis((2,6-dimethyl-
-phenyl)amino)ethane dihydrochloride, 40 mL of triethyl ortho-
3
H
14
D
6
N
2
2
, m/z 352; found, m/z 352.
-phenyl)imidazolinium
(
3
d
3
formate, and 2 drops of 96% formic acid was heated in a distillation
apparatus until the distillation of ethanol ceased. Once the temper-
ature rose toward 130 °C, the heating was stopped and the reaction
mixture was cooled to room temperature. At this point, a white
solid precipitated, which was isolated by vacuum filtration. Re-
crystallization from 2:1 CH
5
1
meta CH, JH,H ) 7.4 Hz), 4.44 (s, 4H, NCH
CN, 75 MHz): δ 161.3, 137.0, 131.3, 130.1, 118.4, 52.1, 17.4 (q,
CD
[
Thermal Decomposition of Methylidene Complex 8′. In the
drybox, a Schlenk tube was charged with 200 mg (0.038 mmol) of
complex 8 dissolved in 10 mL of degassed benzene. The tube was
taken out of the glovebox, connected to argon, and immersed in a
60 °C oil bath. Ethylene gas was bubbled rapidly through the
solution for 30 min, resulting in a color change from pink to dark
brown. The mixture was sampled to confirm full conversion of the
3
CN-ethanol afforded 1.67 g (5.7 mmol,
1
8%) of white crystals. H NMR (CD CN, 300 MHz): δ 9.38 (s,
3
3
H, NCHN), 7.32 (t, 2H, para CH, JH,H ) 7.7 Hz), 7.24 (d, 4H,
2
13
2 3
). C NMR (CD -
5
benzylidene by H NMR.
1
3
, JC,D ) 19.0 Hz). MS (HRMS/EI): calcd for C19
H
11
D
6
2
N Cl
+
M] , m/z 291.2609; found, m/z 291.2621.
The mixture was then stirred at 60 °C for 3 days, after which
the solvent was evaporated in vacuo. The flask was taken back
into the drybox and transferred into a J. Young valve NMR tube.
Isomerization Experiment. Once no more carbene signal was
(j) Synthesis of 1,3-Bis(2,6-dimethyl-d
3
-phenyl)-2-(trichlo-
romethyl)imidazolidine (G). To a solution of chlorine salt E (1.67
g, 5.7 mmol) in dry chloroform was added 0.230 g (5.7 mmol) of
a 60% dispersion of sodium hydride. The reaction mixture was
stirred at room temperature for 2 h and subsequently filtered. The
filtrate was evaporated into a yellowish solid, which was flushed
through a silica plug with 9:1 hexane-ethyl acetate. The ligand
was further purified by recrystallization in boiling hexane, affording
3
3
NMR (CDCl , 300 MHz): δ 7.03 (m, 6H, para, meta CH), 5.66
(
1
visible by H NMR spectroscopy (δ 19.63 and 18.40 ppm for 8
and 8′, respectively), the catalytic mixture was considered fully
decomposed. The J. Young valve NMR tube was fitted onto a
vacuum hookup, and the solvent was evaporated. The tube was
then taken into the drybox, and 0.5 g of 1-octene was added to the
1
decomposed mixture. The tube was allowed to stand at 55 °C for
.28 g (4.7 mmol, 82%) of a white solid (mp 118-120 °C). H
1
2
4 h, and the isomer content was determined by H NMR by
13
integration of the internal olefin peak at δ 5.40 ppm.
s, 1H, CHCCl
3
), 3.98 (m, 2H, NCH
2
), 3.37 (m, 2H, NCH
2
).
C
NMR (CDCl
3
, 75 MHz): δ 144.1, 138.4, 134.2, 129.5, 129.3, 125.6,
5
1
07.9, 86.1, 51.8, 19.0 (q, CD
calcd for C19
91.2622.
k) Synthesis of Complex 8. The procedure was adapted from
3
, JC,D ) 19.0 Hz). MS (HRMS/EI):
Acknowledgment. We thank the National Science Founda-
tion and the Army Research Office for generous funding of this
work. K.A.A. also acknowledges the University of Florida for
funding of the purchase of the X-ray equipment.
+
H
10
D N
12 2
[M + H] , m/z 291.2609; found, m/z
2
(
52
the literature. In the drybox, a flame-dried Schlenk flask was
charged with 0.879 g (1.07 mmol) of catalyst [Ru]1 and 1.640 g
Supporting Information Available: Tables giving full crystal-
lographic data for complex 8. This material is available free of
charge via the Internet at http://pubs.acs.org.
(
52) Trnka, T. M.; Morgan, J. P.; Sanford, M. S.; Wilhem, T. E.; Scholl,
M.; Choi, T.-L.; Ding, S.; Day, M. D.; Grubbs, R. H. J. Am. Chem. Soc.
003, 125, 2546-2558.
2
OM0509894