Bimolecular Coupling as a Vector for Decomposition of Fast-Initiating Olefin Metathesis Catalysts

Article abstract of DOI:10.1021/jacs.8b02709

The correlation between rapid initiation and rapid decomposition in olefin metathesis is probed for a series of fast-initiating, phosphine-free Ru catalysts: the Hoveyda catalyst HII, RuCl₂(L)(=CHC₆H₄-o-OⁱPr); the Grela catalyst nG (a derivative of HII with a nitro group para to OⁱPr); the Piers catalyst PII, [RuCl₂(L)(=CHPCy₃)]OTf; the third-generation Grubbs catalyst GIII, RuCl₂(L)(py)₂(=CHPh); and dianiline catalyst DA, RuCl₂(L)(o-dianiline)(=CHPh), in all of which L = H₂IMes = N,N′-bis(mesityl)imidazolin-2-ylidene. Prior studies of ethylene metathesis have established that various Ru metathesis catalysts can decompose by β-elimination of propene from the metallacyclobutane intermediate RuCl₂(H₂IMes)(κ²-C₃H₆), Ru-2. The present work demonstrates that in metathesis of terminal olefins, β-elimination yields only ca. 25-40% propenes for HII, nG, PII, or DA, and none for GIII. The discrepancy is attributed to competing decomposition via bimolecular coupling of methylidene intermediate RuCl₂(H₂IMes)(=CH₂), Ru-1. Direct evidence for methylidene coupling is presented, via the controlled decomposition of transiently stabilized adducts of Ru-1, RuCl₂(H₂IMes)L_n(=CH₂) (L_n = py_n′; n′ = 1, 2, or o-dianiline). These adducts were synthesized by treating in situ-generated metallacyclobutane Ru-2 with pyridine or o-dianiline, and were isolated by precipitating at low temperature (-116 or -78 °C, respectively). On warming, both undergo methylidene coupling, liberating ethylene and forming RuCl₂(H₂IMes)L_n. A mechanism is proposed based on kinetic studies and molecular-level computational analysis. Bimolecular coupling emerges as an important contributor to the instability of Ru-1, and a potentially major pathway for decomposition of fast-initiating, phosphine-free metathesis catalysts.

Full text of DOI:10.1021/jacs.8b02709

Subscriber access provided by UNIV OF DURHAM
Bimolecular Coupling as a Vector for Decomposition
of Fast-Initiating Olefin Metathesis Catalysts
Gwendolyn A Bailey, Marco Foscato, Carolyn S. Higman,
Craig S. Day, Vidar R. Jensen, and Deryn Elizabeth Fogg
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b02709 • Publication Date (Web): 13 Apr 2018
Downloaded from http://pubs.acs.org on April 13, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 18
Journal of the American Chemical Society
1
2
3
4
5
6
Bimolecular Coupling as a Vector for Decomposition of Fast-
Initiating Olefin Metathesis Catalysts
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
†
‡
†
†
‡
Gwendolyn A. Bailey, Marco Foscato, Carolyn S. Higman, Craig S. Day, Vidar R. Jensen,
,
†
Deryn E. Fogg*
†Center for Catalysis Research and Innovation, and Department of Chemistry and Biomolecular Sciences, Univer-
‡
sity of Ottawa, Ottawa, ON, Canada K1N 6N5. Department of Chemistry, University of Bergen, Allégaten 41, N-
5
007 Bergen, Norway
KEYWORDS: Olefin metathesis, catalyst decomposition, bimolecular coupling, catalyst initiation, high activity
ABSTRACT: The correlation between rapid initiation and rapid decomposition in olefin metathesis is probed for a
i
series of fast-initiating Ru catalysts: the Hoveyda catalyst HII, RuCl
2
(L)(=CHC
6
H
4
-o-O Pr); the Grela catalyst nG (a de-
i
rivative of HII with a nitro group para to O Pr); the Piers catalyst PII, [RuCl
2
(L)(=CHPCy
3
)]OTf; the third-generation
IMes =
Grubbs catalyst GIII, RuCl
2
(L)(py)
2
(=CHPh); and dianiline catalyst DA, RuCl
2
(L)(o-dianiline)(=CHPh) (L = H
2
N,N’-bis (mesityl)-imidazolin-2-ylidene). Prior studies of ethylene metathesis established that various Ru metathesis
2
catalysts can decompose by β-elimination of propene from metallacyclobutane RuCl
2
(H
2
IMes)(κ -C
3
H6) Ru-2. The pre-
sent work demonstrates that in metathesis of terminal olefins, β-elimination yields only ca. 25–40% propenes for HII,
nG, PII or DA, and none for GIII. The discrepancy is attributed to competing decomposition via bimolecular coupling
of methylidene intermediate RuCl
2
(H
2
IMes)(=CH
2
) Ru-1. Direct evidence for methylidene coupling is presented, via the
(H IMes)L (=CH ) (L = pyn’; n’ = 1, 2, or o-
controlled decomposition of transiently-stabilized adducts of Ru-1, RuCl
2
2
n
2
n
dianiline). These adducts were synthesized by treating in situ-generated metallacyclobutane Ru-2 with pyridine or o-
dianiline, and isolated at low temperature (–116 °C or –78 °C, respectively). On warming, both undergo methylidene
coupling, liberating ethylene and forming RuCl
2
(H
2
IMes)L . A mechanism is proposed based on kinetic studies and
n
molecular-level computational analysis. Bimolecular coupling emerges as an important contributor to the instability of
Ru-1, and a potentially major pathway for decomposition of fast-initiating, phosphine-free metathesis catalysts.
To aid in systematic analysis, the classification scheme in
INTRODUCTION
Chart 2 is proposed. Class A metathesis catalysts, exem-
Olefin metathesis represents an exceptionally powerful,
general methodology for the catalytic assembly of car-
plified by the second-generation Grubbs catalyst GII,
initiate slowly, but the ligand dissociated from the
precatalyst is slow to recapture the active species Ru-1.3
1
bon-carbon bonds. Ru-catalyzed olefin metathesis, long
embraced in academia, is now beginning to see industri-
Class B catalysts (e.g. HII or the recently-reported diani-
4
2
al uptake. Reports from pharmaceutical manufacturing,
line catalyst DA) initiate readily, but recapture of the
however, highlight catalyst productivity as a challenge
active species is facile, resulting in rapid shuttling into
and out of the catalytic cycle.3,5 Class C catalysts (e.g. the
2
in process chemistry. Improved understanding of cata-
lyst decomposition, particularly the pathways operative
for the most vulnerable active species (Chart 1), is critical
to guide process implementation and catalyst redesign.
Chart 1. Key Active Species in Olefin Metathesis.
third-generation Grubbs catalyst GIII, the Grela catalyst
nG, or the Piers catalyst PII) also initiate rapidly, and
recapture of Ru-1 by the released ligand is minimal.3,6
The trade-off between activity and stability evident from
7
recent prominent reviews underscores the importance
of understanding the decomposition pathways for fast-
initiating Class B/C metathesis catalysts, in particular.
Chart 2. Classification of Metathesis Catalysts.
The ease with which catalysts enter and exit the active
cycle is fundamental to their productivity and stability.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 18
Bespalova and co-workers likewise invoked β-
elimination / reductive elimination from Ru-2 to account
Class A
Cl
Class B
Ph
Cl
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Ph
Cl
IMes
Cl
H
2
Ru
NH2
H IMes Ru PCy
2
H IMes Ru
OPr
i
for formation of propene upon heating HII in the pres-
ence of ethylene.12,13 Propene is thus a key marker for the
β-elimination pathway in these experiments. (Caution
must be exercised, however, to ensure that no Brønsted
base is present that can generate propenes via an alterna-
tive process, which commences with metallacyclobutane
deprotonation).14,15
2
3
GII Cl
HII Cl
DA H N
2
NO
2
Class C
PCy
3
Ph
Cl
IMes
Cl
Cl
H
2
Ru
PII
OTf
Cl
2
H IMes Ru py
i
H
2
IMes Ru
OPr
GIII
Cl
nG Cl
py
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
A related pathway was proposed based on density func-
tional theory (DFT) calculations, for terminal olefins
bearing a suitable aliphatic substituent.16 For substrates
in which an exocyclic β-H is present, expansion of the
metallacyclobutane ring was predicted, with ultimate
loss of this ligand as an olefin (Scheme 1c).
Intrinsic Decomposition Pathways Established for the
Dominant Ru Metathesis Catalysts. The decomposition
chemistry intrinsic to the PCy -stabilized catalysts of
3
Class A has been extensively studied. GII, for example,
although robust as the precatalyst, readily decomposes
once converted into the metathesis-active methylidene
intermediate Ru-1. Methylidene abstraction from Ru-1
by free PCy
(Scheme 1a) is widely documented,8,9 alt-
hough nucleophilic primary amines such as NH n
compete to abstract this key ligand.9d For phosphine-free
Class B/C catalysts such as HII and PII, the sole intrinsic
decomposition pathway for which experimental evi-
dence has been reported to date is depicted in Scheme
In comparison, the decomposition of pyridine-stabilized
catalysts such as GIII is poorly understood. Sponsler
and coworkers reported that GIIIBr (in which the pyri-
3
2
Bu can
dine ligands are 3-bromopyridine) decomposed within
seconds on exposure to ethylene.17 Hong and Grubbs8a
undertook the corresponding reaction of GIII with eth-
ylene, with the intention of synthesizing methylidene
species GIIIm (Scheme 2). The latter complex, a pyridine
1
b. β-Hydride elimination from the metallacyclobutane
analogue of GIIm RuCl
2
(H
2
IMes)(PCy
3
)(=CH ) (the rest-
2
intermediate Ru-2 generates an allyl hydride, from
which propene is liberated by reductive elimination.
Evidence for this pathway was established by the Piers
group, via synthesis of an isotopologue of Ru-2 bearing
ing state species formed in metathesis by GII), was too
unstable even to observe in situ at RT. Instead, the sole
product identified was the tris-pyridine derivative Ru-3.
(Ru-3 was also formed, and crystallographically charac-
1
3
a C-labelled metallacyclobutane ring. Decomposition of
the labelled complex afforded 13C
, unequivocally
terized, on decomposition of GIIm in the presence of
pyridine).8a The absence of [Me-py]Cl in these reactions
3
H
6
confirming the origin of the propene byproduct in the
metallacyclobutane ring.10 Of note, Eisenstein and co-
workers reported that β-H transfer within the metallacy-
clobutane intermediate is also the key initial step in de-
was explicitly noted, indicating that pyridine – unlike
or primary amines – is insufficiently nucleophilic18
PCy
3
(or insufficiently basic) to attack the methylidene ligand.
The fate of the methylidene ligand was not determined.
Scheme 2. Reported Decomposition Behavior of GIII.a
0
olefin metathesis catalysts.11
composition of d ML
Scheme 1. Intrinsic Decomposition Pathways for Cata-
lysts of Classes A–C (where L = H
IMes).a
4
2
Ph
C2H4
1 atm)
Cl
H IMes Ru
expected:
not observed
Cl
(
py
GIIIm py Cl
H IMes Ru
py
Cl
2
2
PCy
3
(
a)
GIII
Cl
Cl
py
PCy
3
[
MePCy ]Cl
py
Cl
GII
L
L
Ru
L
3
Ru
Cl
+
ruthenium
products
Ru-1 Cl
Cl
H IMes Ru py
Me
N
2
Ru-3
Cl
(
b)
Cl
Cl H
py
not observed
+ ruthenium products
β-elimination
β-elimination
PII
fate of methylidene
ligand unknown
L
Ru
Ru
Cl
or HII
+ ruthenium
products
Cl
aLoss of [Me-py]+ via deprotonation (cyclometallation) of
IMes was postulated: cf. footnote a in Scheme 1a.
Ru-2
H2
Cl
Cl H
Ru
(
c)
Proposed Role for Bimolecular Coupling. We speculat-
ed that bimolecular coupling of the 14-electron interme-
diate Ru-1, with elimination of the methylidene ligands
as ethylene, might contribute to decomposition for all of
the fast-initiating Ru metathesis catalysts. This pathway
would go unrecognized during the “ethenolysis” exper-
iments described above, because ethylene (the marker
for methylidene coupling) is masked by the reagent gas.
HII
L
Ru
L
+
ruthenium
products
Cl
Cl
a
The proton required to eliminate the σ-alkyl ligand in path
IMes ligand.9a
(
a) is supplied by cyclometallation of the H
2
ACS Paragon Plus Environment

Page 3 of 18
Journal of the American Chemical Society
of propene in these experiments provided important
By the same token, any methylidene coupling during
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
metathesis of 1-olefins – the majority of substrates –
would be masked by ethylene formed as the coproduct
of metathesis.
qualitative evidence for β-elimination from the metal-
lacyclobutane Ru-2 (Scheme 1b). However, the yield of
propene based on Ru, and hence the extent of this path-
way, was not determined.
Bimolecular coupling has been extensively documented
0
for d metathesis catalysts, and is indeed well estab-
We therefore began by seeking to quantify the propene
byproducts generated during styrene metathesis. Here
propene and β-methylstyrene (Figure 1) serve as mark-
ers for catalyst decomposition via β-elimination from
Ru-2 or a Ph-substituted metallacyclobutane, respective-
ly. This experiment serves two purposes. First, it reports
on the importance of this decomposition route. Second-
ly, it shifts the focus to 1-olefins, a family of metathesis
substrates of very broad relevance. Styrene is chosen
because, unlike most 1-olefins, it cannot isomerize. This
is critical to prevent formation of “false” propenyl
markers: that is, propenes formed by isomerization-
metathesis (see SI, Section S1), rather than β-elimination.
Accordingly, metathesis of styrene was undertaken with
HII, nG, GIII, PII, and DA (1 mol%; Figure 1). These
experiments were carried out in NMR tubes completely
lished across group 3-7 chemistry.7,19 The rate of decom-
position is, unsurprisingly, sensitive to the bulk of R in
the [M]=CHR species: accordingly, it is generally rapid
for methylidene species, and much slower for complexes
containing bulky alkylidenes. In some cases, the eth-
ylene liberated by coupling of methylidene complexes
afforded isolable ethylene adducts, an important aid in
confirming the operation of bimolecular coupling.7,19
Although bimolecular coupling is also widely accepted
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
for “first-generation” Ru catalysts, it is regarded as con-
siderably less likely for the important Ru-NHC catalysts,
owing in part to the steric impediment to approach of
two RuCl
inhibited for GII and its resting-state species GIIm, for
which slow loss of PCy limits the concentration of Ru-1
present at any given time. Rapid olefin binding to Ru-1
a distinct feature of the NHC catalysts, which led Chen
2
(NHC)(=CH
2
) molecules.20 It is undoubtedly
3
22
filled with solvent, to minimize loss of volatile propene
to the headspace. This results in co-retention of ethylene,
which was expected to maximize formation of metal-
lacyclobutane Ru-2, and consequently propene elimina-
tion. Notably, however, no propenes were observed for
GIII, and <40% propenes for HII, nG, PII, and DA.
Clearly, some additional decomposition pathway is
operative. If bimolecular coupling indeed accounts for
the balance, it is a much more significant contributor to
decomposition of Class B/C catalysts than has been con-
sidered to date.
(
to term them “high commitment”,21 and Piers “ole-
finophilic”10b) further limits the concentration of Ru-1.
Finally, the kinetically dominant decomposition path-
way for GII during metathesis is typically abstraction of
the methylidene ligand by free PCy
Scheme 1a above.8,9
3
, as shown in
For phosphine-free HII, bimolecular coupling is thought
to be limited by facile “boomerang” recapture of Ru-1 by
5
free isopropoxystyrene. This proposition is difficult to
examine, however. HII itself is sterically protected
against such coupling, while the active methylidene
intermediate Ru-1 is spectroscopically unobservable.
Important alternative opportunities for insight are of-
fered by the o-dianiline catalyst DA and pyridine cata-
lyst GIII. We considered that the labile N-donor ligands
in these precatalysts could offer potential access to tran-
siently-stabilized methylidene species (that is, adducts of
Ru-1), if suitable synthetic routes could be envisaged.
Here we report the successful low-temperature synthesis
of such adducts, and the first direct evidence that bimo-
lecular coupling, with loss of the methylidene ligand as
ethylene, represents a major pathway for decomposition
of Class B/C metathesis catalysts. Further, we demon-
strate that the contribution of this pathway to decompo-
sition of HII, the dominant Class B catalyst in current
use, has almost undoubtedly been underestimated.
100
2
Ph
propenes
R
1
mol % Ru
RT, I.S.
β-elimination, then
reductive elimination
Ph
+
Ph
39
(R = Ph, H)
30
27
25
0
metathesis
0
HII nG PII
DA GIII
Figure 1. Quantifying the β-Elimination Pathway in De-
composition of Fast-Initiating Metathesis Catalysts: Propene
Yields at Full Catalyst Decomposition. (Yields based on Ru
precatalysts; reactions in C
6
D
6
except for PII, for which
). I.S. = internal standard.
solubility required use of CD
See also Scheme 1b.
2
Cl
2
Examining Alternative Decomposition Pathways: In-
sight from the Nature of the Ru Products. The py lig-
ands in GIII offer opportunities to trap the Ru products
of decomposition. Of particular interest is the fate of the
RESULTS AND DISCUSSION
Quantifying Decomposition via β-Elimination. Prior
studies of the decomposition of phosphine-free rutheni-
um metathesis catalysts, as noted above, focused on the
behavior of PII and HII under ethylene. The observation
H2IMes ligand in these products, as NHC activation
and/or cyclometalation are common features in numer-
ous potential pathways, including Buchner expansion
7
(see below). We therefore sought to identify the methyl-
ACS Paragon Plus Environment

Journal of the American Chemical Society
idene-free Ru species formed on reaction of GIII with
Page 4 of 18
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
styrene. Scheme 3 depicts the major products observed
under conditions corresponding to those of Figure 1: that
is, the known tris-pyridine complex Ru-3 (a known
thermodynamic sink in this chemistry; see above),8a its
bis-pyridine analog Ru-3’, and a new species assigned as
ethylene adduct Ru-4.
Scheme 3. Bimolecular Decomposition of GIII During
Olefin Metathesis in a Sealed System.a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Ph
Ph
Cl
Cl
RT, 0.5 h
H
2
IMes Ru
py
Cl
H IMes Ru
2
+
I.S., C D6 Ph
6
Ru-1
A
Cl
GIII py
–
2 py
+
+
100
Ph
+ Ru-1
metathesis products
py
py
Cl
py
Cl
Cl
py
Ru-3
(98%)
L
Ru py
+
L
Ru py
+
L
Ru
Cl
Cl
Cl
Ru-3
Ru-3’
Ru-4
(19%)
py
(45%)
(14%)
ruthenium products observed; L = H IMes
2
a
Ethylene is generated by both metathesis of styrene, and
methylidene coupling.
These sealed-tube reactions generate Ru-3/3’ in ca. 60%
yield, and Ru-4 in ca. 20% yield based on ruthenium
(
Figure 2a; inverted spectrum). The presence of an eth-
Figure 2. Ruthenium Decomposition Products: Identifica-
tion and Mechanistic Implications. (a) 1H NMR spectra
ylene ligand in Ru-4 is supported by the observed trans-
formation of this complex into Ru-3’ when the solution
is degassed to remove ethylene, and its reappearance
when ethylene is reintroduced (Figure S13). Labile coor-
corresponding to Scheme 3: diagnostic py o-CH region
(C
6
D6, 300 MHz). Upper trace: Ru-3, formed by adding
pyridine to (inverted trace) sample at full decomposition.
For full spectra, see Fig. S13. (b) Decomposition pathways
ruled out by quantitative formation of Ru-3.
dination of C
2
H is common in electron-rich Ru com-
4
plexes.8a,14,23 As further evidence for a weakly-bound
ethylene ligand in Ru-4, the latter complex undergoes
conversion to tris-py complex Ru-3 when pyridine is
added (as does bis-py complex Ru-3’). As shown in the
upper NMR trace of Figure 2a, Ru-3 is then the sole ob-
servable Ru species, being present in 98% yield based on
starting GIII. As an indicator of generality, it should be
noted that Ru-3 is likewise observed for HII on succes-
sive treatment with styrene and pyridine, and (as dis-
cussed below) on decomposition of the methylidene and
ethylidene complexes GIIIm and GIIIe.
A key structural feature in Ru-3 is the intact H
ligand, which rules out decomposition processes that
involve activation of H IMes. This precludes, for exam-
2
IMes
2
ple, base-induced deprotonation of the metallacyclobu-
tane, which would generate a Ru dimer containing a
cyclometallated H
2
IMes ligand14 (Figure 2b, left) as well
as “false” propenyl markers, as discussed above. The
24
poor Brønsted basicity of pyridine is a key parameter in
inhibiting this pathway.
Also ruled out is pyridine-induced attack of the
[Ru]=CHR carbon on a mesityl ring, which would give a
cycloheptatriene product following Buchner expansion
Figure 2b, right).25,26 Such transformations of GII, HII,
(
and other Ru-H IMes catalysts were deliberately in-
2
duced by Diver and co-workers, via reactions with CO
or isonitriles.25 Coordination of these π-acids heightens
the electrophilicity of the Ru=CHR carbon, promoting
cyclopropanation of a mesityl ring and ensuing ring
expansion.25,26a Computational analysis by the Cavallo
group suggested that pyridines are insufficiently π-
acidic to cause Buchner expansion of benzylidene com-
ACS Paragon Plus Environment

Page 5 of 18
1
Journal of the American Chemical Society
plexes,26 and the present studies bear out this predic-
Ph
Me
Cl
Ph
Cl
tion.27
(1 atm)
H
2
IMes Ru
py
Cl
+
H IMes Ru
2
py
Cl
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
6
°C, C H6
Evidence for Bimolecular Coupling of [Ru]=CHR (R =
Ph, Me). The foregoing demonstrates that several de-
composition pathways known from other contexts are
either inoperative, or less important than hitherto pre-
sumed. We next turned to establishing whether bimo-
lecular coupling is operative. Unambiguous evidence for
such coupling is seen for the benzylidene complex GIII.
This precatalyst (in fact, typically a mixture of GIII and
its mono-pyridine derivative GIII’; see below) eliminates
stilbene A in up to 75% yield on prolonged heating at 60
°C (Scheme 4). The major Ru products are again Ru-3
and its bis-pyridine analog Ru-3’, formed in up to 70%
yield. Batch-to-batch variations in the proportions of
these products (e.g., 1.5:1 to the reverse ratio) are not
6
GIII
py
10 min
GIIIe
py
93% (isolated)
observed decomposition products
py
py
C D
6 6
Cl
Cl
60 °C
H IMes Ru py + H IMes Ru py
25 min
2
2
Ru-3
30%)
Cl
Ru-3’
(48%)
Cl
py
(
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
+
+
+
(24%)
(24%)
(45%)
Evidence for Bimolecular Coupling of [Ru]=CH . While
2
the observation of alkylidene coupling products in up to
75% yield for the benzylidene and ethylidene complexes
is suggestive, the sterically unprotected methylidene
ligand could be subject to additional reaction pathways.
To clarify the decomposition behavior of methylidene
intermediate Ru-1, we sought a weakly stabilized adduct
that would permit direct study. No such complex is
described in the literature. The sole isolable, metathesis-
active Ru methylidene complexes reported to date are
the “Grubbs methylidenes”, i.e. GIIm8,31 and its first-
unexpected, given the known variability in the number
of pyridine ligands present in the precatalyst28,29 (a GIII–
GIII’ mixture, although exchange averaging results in
1
observation of a single benzylidene peak by H NMR
analysis). Also observed, albeit as a minor product (in-
variably <5%), is RuCl
2
(py)
4
Ru-5. Formation of the latter
IMes loss via a minor
NHC-free complex indicates H
2
additional process as yet undetermined.30
generation analog RuCl
these complexes to probe bimolecular coupling is pre-
cluded by the low lability of the PCy
ligand33 (which
2
(PCy
3
)
2
(=CH
), GIm.32a Use of
2
Scheme 4. Representative Product Speciation on Ther-
mal Decomposition of GIII.
3
limits the concentration of Ru-1 present at any given
Ph
time), and by facile abstraction of the methylidene lig-
Cl
Ph
6
0 °C, 5 d
]Cl,8,9 as cited in the Introduction. As
and as [MePCy
3
H IMes Ru
py
Cl
+
ruthenium
products
2
I.S., C D
6
6
noted above, Hong and Grubbs attempted to synthesize
GIII
Ph
A
py
(74%)
the corresponding pyridine-stabilized species GIIIm via
reaction of GIII with ethylene, but were unable to ob-
serve GIIIm even in situ.8a Few other Ru methylidene
py
py
py
Cl
Cl
Cl
H
2
IMes Ru py +
H
2
IMes Ru py
+
py Ru py
complexes are known, and these rare examples are not
metathesis-active.32
Cl
Cl
Cl
Ru-3
34%)
Ru-3’
(28%)
Ru-5
(<2%)
py
py
(
Nevertheless, we envisaged that Ru-1 might be success-
fully trapped by synthesizing the metallacyclobutane
complex Ru-2 in situ,10 and introducing a stabilizing
ancillary ligand to trigger retro-addition. While pyridine
is an obvious candidate ligand, the literature reports
emphasize its limited stabilizing ability.8a,17 o-Dianiline
ruthenium products observed
Ethylidene complex GIIIe (Scheme 5) decomposes dra-
matically faster than GIII. We were able to isolate GIIIe
by stirring GIII under 1 atm cis-2-butene at the mini-
mum accessible solution temperature in benzene, then
lyophilizing the solvent to prevent premature decompo-
(Scheme 6) offers an attractive alternative. This ancillary
sition upon concentrating. Heating a C
6
D6 solution of
ligand, originally chosen to maximize catalyst productiv-
GIIIe and an internal standard at 60 °C causes loss of the
4
ity in ring-closing macrocyclization, is similarly well-
alkylidene signal over 25 min, with evolution of up to
suited to stabilization of Ru-1. Design criteria common
to both of these objectives include low nucleophilicity
and low Brønsted basicity (essential to prevent ligand-
mediated methylidene or proton abstraction, respective-
4
5% 2-butene. Also observed are propene and 2-pentene
(
24% each), indicating isomerization–metathesis of 2-
butene. On adding pyridine, the Ru products again con-
vert into Ru-3 (85% yield, based on starting GIIIe).
Scheme 5. Synthesis and Bimolecular Decomposition
of GIIIe.
9
ly). Also critical is a balance between the coordinating
properties required to isolate an adduct of Ru-1, vs. the
lability required to readily release Ru-1. These conflicting
demands are reconciled via a combination of good lig-
and donor ability (achieved, for o-dianiline, via a combi-
nation of donicity, steric accessibility at the nitrogen
34
sites, and flexible chelation), and high (hemi)lability.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 18
are highlighted with bars that approximate the colours of
the complexes.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 6. Synthesis of o-Dianiline Adduct Ru-6.a
The corresponding experiments with GIIIm, synthe-
sized by adding 2 equiv py (Scheme 7), afforded a mix-
ture of GIIIm and its mono-pyridine derivative GIIIm’.
These species were generated in a 30:70 ratio, as deter-
mined by integration of the two distinct methylidene
signals (19.35 and 18.63 ppm, respectively) at –50 °C. The
higher lability of the monodentate pyridine ligand ren-
ders these complexes much more unstable than Ru-6,
and considerably more challenging to isolate. However,
they were successfully obtained by carrying out pyridine
addition at –78 °C to form GIIIm/m’, and precipitating
the product mixture by cannula addition of cold pentane
at ca. –120 °C. Even under these conditions, competing
decomposition (loss of ethylene) was evident. Nonethe-
less, the methylidene products were isolated as a yellow
solid, in ca. 60% yield based on starting PII, by decant-
ing the pentane and drying in vacuo. Their stability was
insufficient to withstand washing, even in the solid state.
o-dianiline
PCy3
NH2
• low nucleophilicity
Cl
IMes
•
•
•
low Brønsted basicity
good Lewis basicity
flexible chelation
H
2
Ru
OTf
PII Cl
H
2
N
• high (hemi)lability
2 2
CD Cl
(1 atm) –50 °C
Cl
Cl
H IMes
1
.1 o-dianiline
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
H IMes
2
Ru
Cl
2
Ru
NH2
–50 °C
Cl
Ru-2
H N
2
PCy3 OTf
Ru-6
+
a
Isolated by precipitating with pentane at –78 °C.
Synthesis of Transiently-Stabilized Methylidene Adducts of
Ru-1. In initial NMR-tube experiments (Scheme 6), a
solution of PII in CD
2
Cl2 was freeze-thaw degassed,
thawed at –50 °C, and exposed to ethylene. A color
change from brown to dark pink occurred within 5 min,
with clean conversion to Ru-2 (Figure 3, top trace). Slow
a
Scheme 7. Synthesis of Pyridine-Stabilized GIIIm/m’.
injection of o-dianiline in CD
2
Cl2 caused a further color
change to green over the next 0.5 h, with loss of the di-
Cl
PCy3
Cl
Cl
(1 atm)
CD Cl
2 py
agnostic upfield signal for the metallacycle (H ; –2.6
β
L
Ru
Cl
L
Ru
py
Cl
L
Ru
OTf
ppm), and emergence of the methylidene singlet for Ru-
2
2
–78 °C
PII Cl
–50 °C Ru-2
(
py)n
6
(19.3 ppm; Figure 3, inverted trace). The diagnostic,
formally diastereotopic NH signals for bound o-
dianiline appear as two broad, unresolved signals at ca.
.6 and 4.3 ppm. Their assignment was confirmed by
injecting D O into the cooled solution, and briefly agitat-
GIIIm/m’
2
a
GIIIm: n = 1; GIIIm’: n = 0; L = H IMes. The 30:70 mixture
2
was precipitated with pentane at –116 °C. The vi-
nylphosphonium coproduct in step 1 is omitted for clarity.
3
2
ing to dissolve the ice. In situ yields of Ru-6 were quanti-
tative. To isolate Ru-6 free of reagent ethylene, the reac-
tion was repeated on 75 mg scale, and Ru-6 was isolated
as a pale green solid in 75% yield by precipitating with a
trickle of cold pentane at –78 °C.
Evidence for Bimolecular Coupling of Transiently-Stabilized
Methylidene Species. To measure the proportion of eth-
ylene released by coupling of dianiline adduct Ru-6, this
complex was dissolved in cold CD
2
Cl2 and added to the
internal standard in a pre-chilled J. Young tube. The
temperature was maintained at –20 °C during sample
preparation by use of a sand-bath chilled in the glovebox
freezer. Warming to RT resulted in 95% decomposition
over 75 min (Figure 4), and formation of methylidene-
free Ru-7 in up to 90% yield. Only 30% yield of ethylene
was detected, however, owing to volatilization into the
headspace. The proportion of ethylene detected in-
creased to ca. 65% when Ru-6 was permitted to decom-
pose in NMR tubes filled to 80% capacity. This
constitutes the minimum headspace volume deemed
safe to prevent explosion on warming (for calculations,
see SI).
1
Figure 3. H NMR spectra showing (top) Ru-2, and (invert-
ed) its o-dianiline-stabilized derivative Ru-6 (300 MHz,
CD
2
Cl2, –50 °C). Diagnostic NMR signals for Ru-2 and Ru-6
ACS Paragon Plus Environment

Page 7 of 18
Journal of the American Chemical Society
R
100
C D , 60 °C
Cl
6
6
1
2
3
4
5
6
7
8
9
H IMes Ru
py
Cl
2
100
CD
2
Cl
RT
2
R
py
Ph GIII
100
99
100
Me GIIIe
30 min
+ I.S.
H
GIIIm
0
0
py
R: Ph
Me
H
R
Cl
+
H
2
IMes Ru py
0
0
0
Ru-3
Cl
R
py
R: Ph
(GIII)
Me
(GIIIe)
H
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(GIIIm)
Figure 5. Bimolecular coupling of [Ru]=CHR: impact of R
on rate of decomposition. Chart shows % alkylidene re-
maining after 30 min at 60 °C, or (inset) at RT, for GIII,
GIIIe, and GIIIm. All complexes shown as bis-py species
for simplicity: in practice, a mixture of mono- and bis-py
complexes is present (of which the latter predominates for
all but GIIIm/m’).
1
Figure 4. Bimolecular coupling of Ru-6: rate curve and H
NMR spectrum (300 MHz, CD
tion.
2
Cl , RT) at 95% decomposi-
2
In these filled-tube experiments, the yield of the Ru de-
composition product Ru-7 dropped to ca. 65%, probably
Mechanism of Bimolecular Coupling. To gain further
insight into the mechanism of bimolecular coupling, we
undertook kinetics and computational analysis.
owing to the improved stability of Ru-C
2
H adducts (cf.
4
Ru-4) at higher solution concentrations of ethylene. New
1H NMR signals were indeed observed in the region
Kinetics studies. Rate experiments focused on dianiline
adduct Ru-6, given its greater ease of handling relative
to GIIIm. At 20 mM Ru-6, decomposition was second-
order in Ru, consistent with rate-limiting bimolecular
coupling of five-coordinate A (Figure 6a). At 1 mM Ru-6,
decomposition is first order in Ru, but the evolution of
ethylene confirms that bimolecular coupling remains
characteristic of the Ru-bound C
2
H
4
ligand (3.5–0.9
ppm),14,23 as well as the NH region (5.0–3.4) for Ru-
dianiline complexes, although the lability of bound eth-
ylene in these complexes precludes isolation.
The corresponding experiments with the pyridine ad-
ducts GIIIm/m’ resulted in complete decomposition
within 20 min, consistent with the high lability of the
pyridine ligands noted. Higher proportions of ethylene
operative. Indeed, the observed yield of C
2
H reached ca.
4
4
0% based on starting Ru-6, despite use of a standard
were detected (76% in CDCl
3
; 70% in CD
2
Cl ), probably
2
NMR-tube headspace to maintain conditions that more
closely approximate bench operations. Volatilization of
ethylene is limited by the small amounts involved (max-
imum theoretical yield 0.5 mM, well below its reported
because mass transfer restrictions in the NMR tube re-
tard partitioning of ethylene into the gas phase over this
short reaction time. Bimolecular coupling is thus ob-
served for both of these exemplary class B/C catalysts,
although the weaker donor ligands characteristic of
Class C accelerate decomposition.
at 296 ±1.5 K).9d
solubility limit of 54 ±3 mM in CD
2
Cl
2
Decomposition at 1 mM is also found to be retarded by
exogenous dianiline. This, and the change in the order of
reaction with respect to [Ru-6], point toward a change in
mechanism when the ruthenium concentration is de-
creased. The rate of coupling of five-coordinate A (path
a) is expected to be highly sensitive to concentration,
given its squared dependence on [Ru-6]. We propose
that at 1 mM Ru-6, loss of dianiline from A (path b) is
faster than coupling of A. Once four-coordinate Ru-1 is
generated, its greatly reduced steric protection relative
to A is expected to result in a significantly faster rate in
Dependence of the Rate of Bimolecular Coupling on
Alkylidene Substituent. The data above support bimo-
lecular decomposition of all [Ru]=CHR complexes exam-
ined, at rates that are qualitatively found to increase as
the bulk of the substituent R decreases. This behaviour,
which parallels trends observed in the early-metal sys-
tems, is shown more explicitly in Figure 5. Thus, after 30
min at 60 °C in C
intact, while its ethylidene or methylidene analogues
GIIIe and GIIIm, respectively) are completely decom-
6
D , benzylidene complex GIII remains
6
(
the bimolecular coupling step (that is, k
4
>> k ).
2
posed. A striking difference between GIIIe and GIIIm is
evident at room temperature, however. After 30 min in
CD
2
Cl , 99% GIIIe remains, while GIIIm/m’ is complete-
2
ly decomposed. The greater resistance to coupling of the
ethylidene complex GIIIe helps account for the widely-
reported35-37 observation of improved metathesis perfor-
mance for 2-methyl olefins, vs 1-olefins.38
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 18
bimolecular decomposition
termediate cannot generate ethylene in a single elemen-
tary step.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(a)
Cl
k1
Cl
IMes Ru
A
k
2
(slow)
H IMes Ru
NH2
H
2
2
k
-1
RDS at 20
mM Ru-6
Cl
Cl
NH
H
2
N
2
H N
2
0.5
Ru-6
+
Ru-7
(
b)
3
k (RDS at
Cl
1
mM Ru-6)
k
4
H IMes Ru
2
metathesis
cycle
k-3
(fast)
Ru-1 Cl
+
H
2
N
NH
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
00
0.5
2
0 mM
23 °C
2nd-order coupling
(at 20 mM)
k
obs = 0.12 M-1 s-1
0
0
0
0
time (h)
1
time (h)
1
1
00
0.2
Figure 7. Computed Gibbs free energy profile (kcal/mol)
along the reaction path involving loss of methylidene from
Ru-1. Energies normalized to o-dianiline adduct Ru-6.
1
1
mM
0 °C
1st-order coupling
(at 1 mM)
Considerably higher in energy than Ru-10 is Ru-11,
precedent for which exists in a crystallographically char-
k
obs = 1.9*10-4 s-1
acterized tungsten methylidene complex that exhibits
reciprocal methylidene-to-tungsten donor interactions.19c
0
-1.4
0
time (h)
2
0
time (h)
2
Shown in the structure of Ru-11 is the corresponding
Figure 6. Top: Proposed mechanism for bimolecular cou-
pling at (a) 20 mM Ru-6; (b) 1 mM Ru-6. (Reactions in
reciprocal Ru=CH →Ru donation. Although the two
2
methylidene units are closer in Ru-11 than they are in
Ru-10, they are geometrically constrained and cannot
easily interact. Attempts to enforce coupling by shorten-
ing the C−C distance to form an ethylene ligand between
the two Ru centers required unacceptably high energies,
CD
2
Cl
2
; H
2
N–NH = o-dianiline). Bottom: Establishing the
2
order of reaction with respect to Ru. To retard reaction and
collect sufficient scans for good signal-to-noise ratios, the 1
mM reaction was conducted at 10 °C.
>
35 kcal/mol higher than Ru-6, and a transition state was
Molecular-Level Computational Studies. DFT analysis fo-
cused on coupling of Ru-1. This species was chosen for
study in light of the kinetics findings above, which point
toward Ru-1 as the key intermediate in decomposition at
catalytically relevant concentrations of Ru-6. In addition,
Ru-1 is common to all of the Class B/C catalysts studied
in this work, including those which (subsequent to initi-
ation) lack a ligand that can stabilize the methylidene
intermediate.
therefore not located.
Much more persuasive as an intermediate on the me-
thylidene-coupling pathway is Ru-8, which is higher in
energy than Ru-10, but lower than Ru-11. In structure
Ru-8, the two Ru=CH units form a cisoid dimer in
2
which the methylidene carbons are sufficiently close and
unconstrained to interact (2.90 Å; for 3D structure, see
Figure S5). Also notable is a striking dissymmetry in the
methylidene bonding interactions. While the calculated
Accordingly, the mechanism by which two molecules of
Ru-1 couple to generate ethylene was examined. The
dimeric structures shown at the left in Figure 7 were
prioritized in light of prior experimental and computa-
Ru=CH
or 1.83 Å, one methylidene ligand interacts with both
metal centers, while the other does not (Ru −C = 2.64 Å;
Ru −C = 3.67 Å; for atom labeling, see Figure 8). Most
2
bond distances are essentially identical, at 1.82
A
B
tional work suggesting their potential importance.6
,16,39,40
B
A
importantly, the orientation of the methylidene groups
of Ru-8 is optimal for C−C bond formation. Less than 2
kcal/mol is needed to reach transition state Ru-9 (the 3D
structure of which is shown in Figure S6), and these two
stationary points have been connected in intrinsic reac-
The energies shown for these structures are normalized
to that of Ru-6. Dimer Ru-10 has multiple precedents in
crystal structures of diruthenium species containing
transoid alkylidene ligands.41 Ru-10 is the lowest-energy
of the intermediates predicted to form on reaction of two
Ru-1 molecules. However, the long H
5.77 Å), and the substantial rearrangement needed for
the two methylidene units to react, imply that this in-
2
C···CH
2
distance
tion coordinate (IRC) calculations. A related Ru
transition state was identified in a recent DFT study
focusing on the formation of alkylidenes
RuCl (H IMes)(=CRR’) in the reaction of RuCl (H IMes)
2
(ꢀ-Cl)
2
(
2
2
2
2
ACS Paragon Plus Environment

Page 9 of 18
Journal of the American Chemical Society
1
6
with olefin. The relatively low energy required to reach
transition state Ru-9 (<20 kcal/mol above Ru-6) under-
scores the ease of bimolecular coupling. Ensuing loss of
ethylene and trapping by free dianiline affords the me-
thylidene-free product Ru-7.
lacyclobutane as central to decomposition. Bimolecular
coupling, in contrast, has been viewed as largely irrele-
vant for the highly active Ru-NHC catalysts. The forego-
ing corrects this perspective. Bimolecular coupling can
now be recognized as an important contributor to de-
composition of fast-initiating ruthenium metathesis
catalysts. Improved activity – in particular, higher initia-
tion efficiency – has long been a major focus of catalyst
design efforts in Ru-catalyzed metathesis. The present
work suggests that such efforts may be undermined by
accelerated decomposition. Further, it highlights the
importance of inhibiting bimolecular elimination of the
methylidene ligand in designing new metathesis cata-
lysts or reaction protocols.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
To clarify the orbital interactions that lead to coupling of
two molecules of Ru-1, we undertook natural bond or-
bital (NBO) analysis42 of intermediate Ru-8 (Figure 8).
This analysis suggests that coupling is enabled by two
principal orbital interactions. First, electron density is
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
donated from the filled π orbital on Ru
ty orbital on Ru (Figure 8a). This interaction establishes
a small, shared electron population between Ru and C ,
and between the two Ru atoms (Wiberg index43 0.14 in
each case; Table S7). Secondly, as shown in Figure 8b,
back-donation of electron density from the filled π or-
bital on Ru
buildup of electron density between C
interaction is strengthened by polarization of the Ru
π bond towards Ru , which results in polarization of the
corresponding π* orbital in the opposite direction.
C→Ru polarization of the alkylidene bond is consistent
B
=C into an emp-
B
A
A
B
EXPERIMENTAL
SECTION
AND
COMPUTATIONAL
A
=C
A
into the π* orbital on Ru
B
=C
B
causes
. This
=C
General Procedures. Reactions were carried out under
in a glovebox or on a Schlenk line. Dichloromethane,
A
and C
B
N
2
B
B
benzene, and hexanes were dried and degassed using a
Glass Contour solvent purification system. Pentane was
B
distilled over MgSO
bridge Isotopes) were degassed by five consecutive
freeze/pump/thaw cycles. CD Cl (Cambridge Isotopes)
was received in sealed ampoules packed under N . All
solvents were stored under N over 4 Å molecular sieves
for at least 16 h prior to use. Styrene (Aldrich, 99%),
methyl 10-undecenoate (Aldrich, 96%), and D O (Cam-
bridge Isotopes) were degassed by five consecutive
freeze/pump/thaw cycles, and stored under N at –35 °C.
4
, then P
2
O
5
. C
6
D6
and CDCl (Cam-
3
with the experimentally-observed electrophilicity of
the [Ru]=CHR carbon.8,9 Previous computational stud-
2
2
4
4
2
ies, as well as the present work (Tables S8, S9), indicate
that this polarization is essentially limited to the π-
component of the bond.
2
2
NBO analysis of the ensuing transition state Ru-9 reveals
the expected enhancement of the two orbital interactions
discussed for Ru-8. Nevertheless, the shared electron
2
Dimethyl terephthalate (DMT: Aldrich, 99%), trimethox-
ybenzene (TMB: Aldrich, 99%), anthracene (Aldrich,
density between the C
berg index43 of 0.27), indicating that most of the C–C
A
and C is low even in Ru-9 (Wi-
B
9
7%), pyridine (anhydrous; Aldrich, 99.8%), ethylene
BOC Ultra-High Purity grade 3.0, 99.9%; Linde), and
cis-2-butene (Lab Network Inc, 99%) were used as re-
bond is formed subsequent to this transition state.
(
ceived.
RuCl (py)
o-Dianiline
([1,1'-biphenyl]-2,2'-diamine)45
2
4
,46 GIII,28 HII,47 DA, and PII (as the trifluo-
4
romethanesulfonate salt)48 were prepared via literature
procedures.
NMR spectra were recorded on an Avance 300, Avance
II 300, or Avance III 600 cryoprobe NMR spectrometer,
at 23 ±2 °C except where otherwise indicated. Chemical
shifts are reported in ppm, and referenced against the
residual proton or carbon signals of the deuterated sol-
1
13
1
vents ( H, C). Overlapping H NMR integrations were
deconvoluted using the “deconvolve and display” func-
tion in Topspin (NMR processing software; v3.5 pl 5).
Oxygen and moisture were excluded from air-sensitive
samples using either screw-capped NMR tubes
Figure 8. Key orbital interactions for Ru-methylidene cou-
pling and C–C bond formation. Dashed lines signify unoc-
cupied orbitals. Charge flow (donation) is indicated by
arrows. Atom labeling in intermediate Ru-8 shown in box.
(
equipped with PTFE septa, where o-dianiline solutions
were to be injected) or 3 mm valved J. Young NMR tubes
Figure S1). Infrared (IR) spectra were recorded on a
CONCLUSIONS
Fast initiation has long been connected to fast decompo-
sition in Ru-catalyzed olefin metathesis. Prior studies of
highly active, fast-initiating Ru-NHC catalysts identified
elimination of propene from the unsubstituted metal-
(
Thermo Scientific Nicolet 6700 Fourier Transform IR
spectrometer equipped with a Smart iTR Attenuated
Total Reflectance (ATR) sampling accessory. GC-MS
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 18
analysis was performed using an Agilent 5975B inert XL styrene (115 ꢀL, 1.0 mmol, 100 equiv) was then added. A
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
EI/Cl instrument equipped with a polysiloxane column.
Anaerobic MALDI mass spectra were collected on a
Bruker UltrafleXtreme MALDI-TOF/TOF mass spec-
trometer interfaced to a glovebox. Samples were cali-
colour change from green to orange was observed with-
in the first 30 min, followed by a change back to green.
No signals for GIII or other [Ru]=CHR species were
1
evident after 35 min: H NMR (C
6
D6, 300 MHz; diagnos-
brated internally using the peaks for pyrene (m/z
tic signals only; Figure S12): δ 19.66 (s, 1H, [Ru]=CHPh of
+
3
4
2
02.0783) and [H
2
IMes•H] (m/z 307.2174).
GIII; none remaining), 9.62 (dt, JHH = 5.2 Hz, JHH = 1.5
Hz, 4H, py o-CH of Ru-3, 45%), 9.25 (br d, 2H, py o-CH
of Ru-4, 19%), 9.16 (br s, 2H, py o-CH of Ru-3’, 14%), 8.19
(s, 2H, CH of anthracene). Adding pyridine (5 ꢀL, 6
equiv) effected immediate conversion to Ru-3 as the sole
py species present (98% yield).
Quantification of propenyl species formed during
styrene metathesis. To maximize retention of propene,
these experiments were conducted in filled J. Young
NMR tubes. A small headspace (0.15 mL; ca. 5% of the
tube volume) was provided to accommodate any unin-
tended thermal expansion of the solvent in these nomi-
nally isothermal (23 °C) experiments.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1H NMR data for the individual complexes are given
below.
In a representative reaction, a capillary tube was inserted
into a J. Young NMR tube as a mechanical aid to mixing.
A solution of HII (24 mg, 0.038 mmol) and anthracene
RuCl
2
(H
2
IMes)(py)
3
Ru-3: 1H NMR (C
3
6
D , 300 MHz;
6
Figure S15): δ 9.62 (dt, JHH = 5.2 Hz, 4JHH = 1.5 Hz, 4H, py
o-CH), 9.39 (dt, JHH = 4.9 Hz, 4JHH = 1.6 Hz, 2H, py o-CH),
6.53 (tt, 3JHH = 7.5 Hz, 4JHH = 1.6 Hz, 1H, py p-CH), 6.41 (tt,
3JHH = 7.4 Hz, 4JHH = 1.5 Hz, 2H, py p-CH), 6.28 (m, 2H, py
m-CH; overlaps with Mes CH), 6.28 (s, 4H, Mes CH;
overlaps with py m-CH), 5.98 (m, 4H, py m-CH), 3.64 (s,
3
(ca. 3 mg; internal standard) in 1.46 mL C
6
D6 was added,
1
and a H NMR spectrum was measured to establish the
starting ratio of HII vs anthracene. Styrene (440 ꢀL, 3.8
mmol, 100 equiv) was then added, and the tube was
inverted several times to mix the solution. Subsequent
controlled mixing was accomplished by attaching the
tube to a rotary evaporator, and setting it to rotate at ca.
15 rpm (Figure S2). Falling of the internal capillary with
every inversion of the tube effects equilibration, and
enables accurate, reproducible quantitation. A color
change from green to brown occurred, followed by pre-
cipitation of a dark green solid over 1 h. Complete loss
4H, NCH
ues in CD
ported in the same solvent.8a
2
), 2.77 (s, 12H, o-CH
3
), 1.88 (s, 6H, p-CH ). Val-
3
2
Cl2 are in excellent agreement with those re-
1
RuCl
2
(H
2
IMes)(py)
2
Ru-3’: H NMR (C
6
D6, 300 MHz;
Figures S15): δ 9.42 (br s, 2H, py o-CH; overlaps with py
o-CH of Ru-3), 9.15 (br s, 2H, py o-CH), 6.67–6.35 (br s,
6H, Mes CH and py CH; overlaps with py m/p-CH of Ru-
3), 6.28 (s, 2H, Mes CH; overlaps with Mes CH of Ru-3),
6.03 (br s, 2H, py m-CH; overlaps with py m-CH of Ru-
1
of HII was evident at this point. H NMR (C
MHz; diagnostic signals only; Figure S11): δ 16.72 (s, 1H,
Ru]=CHAr of HII; none remaining), 8.19 (s, 2H, Ar CH
of anthracene), 7.00 (s, 2H, =CH of stilbene), 6.58 (dd, 3JHH
6
D
6
, 300
3), 3.48 (br s, 4H, NCH
2
), 2.75 (br s, 12H, o-CH
3
), 1.99 (br
1
[
s, 6H, p-CH
3
). H NMR (CD
2
Cl2, 300 MHz): δ 8.94 (br s,
2H, py o-CH), 8.60 (br s, 2H, py o-CH), 7.51 (br s, 1H, py
p-CH), 7.25 (s, 1H, py p-CH), 7.00 (br s, 2H, py m-CH),
6.59 (br s, 2H, py m-CH), 6.34 (s, 4H, Mes CH), 3.97 (s,
=
1
18 Hz, 3JHH = 11 Hz, =CHPh of styrene), 6.19 (dt, 3JHH
Hz, 3JHH
Hz, 1H, =CHCH Ph of 1,3-
=
6
=
7
2
diphenylpropene;15 not observed), 6.03 (dq, 3JHH = 15.9
4H, NCH
2
), 2.48 (s, 12H, o-CH
3
), 2.15 (s, 6H, p-CH ). All
3
Hz, 3JHH = 6.8 Hz, 1H, =CHCH
2%), 5.27 (s, C
propene =CH
pattern;49 the remaining multiplet is par-
3
of β-methylstyrene;
15
1
H NMR signals for Ru-3’ are broad (ω1/2 ~5–50 Hz) at
RT, indicating fluxionality. A monomeric formulation is
supported by DOSY-NMR analysis, which shows slight-
ly faster diffusion for Ru-3’ than Ru-3. Key 13C NMR
1
2
H ), 5.01–4.92 (m, three-quarters of the
4
2
tially obscured by signal for excess styrene; 18%). The
1
13
corresponding propenes arising from elimination of the
signals (located by H– C HMQC; C
(py o-CH), 152.6 (py o-CH), 121.7 (py m-CH), 52.4
(NCH ), 20.6 (Mes p-CH ), 19.2 (Mes o-CH ).
RuCl(H D , 300
6
D , 300 MHz): 154.0
6
isopropoxyphenyl-substituted
metallacyclobutane15
were not observed, perhaps reflecting the large excess of
styrene present in solution. Adding 4 equiv pyridine to
the reaction results in slow dissolution of the green solid,
and a colour change to from green to orange. After 24 h:
2
3
3
2
IMes)(C
2
H
4
)(py) Ru-4: 1H NMR (C
6
6
3
MHz; Figure S12): δ 9.30 (br d, JHH = 5.3 Hz, 2H, py o-
CH), 6.63 (py m-CH; overlaps with styrene =CH; 2D-
1H NMR (C
dt, 3JHH = 5.2 Hz, 4JHH = 1.5 Hz, 4H, py o-CH of Ru-3,
8%).
6
D6
, 300 MHz; diagnostic signals only): δ 9.62
detected via COSY), 3.59 (br s, 4H, NCH
Mes CH ), 2.67 (br s, 3H, Mes CH ), 2.49 (br s, 6H, Mes
CH ), 2.11 (br s, 3H, Mes CH ), 1.70 (br s, 3H, Mes CH ).
are masked by
2
), 3.05 (br s, 3H,
(
3
3
7
3
3
3
Identification of Ru species formed on metathesis of
styrene by GIII. In a representative experiment, a solu-
tion of GIII (7.3 mg, 0.010 mmol) and anthracene (ca. 1
Signals for Mes CH, py p-CH, and C
2
H
4
1
1
overlap with Ru-3/3’ and styrene. H– H NOESY: δ 9.30
(py o-CH) and 6.63 (py m-CH), 9.30 and 2.67 (Mes CH
Stirring a mixture of Ru-3/3’/4 in C at RT under N for
3 h resulted in a colour change from green to orange,
and complete consumption of Ru-4 (Figure S13). At t :
3
).
mg; internal standard) in 385 ꢀL C
Young NMR tube. A H NMR spectrum was measured
6
D
6
was added to a J.
6
D6
2
1
to establish the starting ratio of GIII vs anthracene, and
0
ACS Paragon Plus Environment

Page 11 of 18
Journal of the American Chemical Society
4
6
5% Ru-3, 19% Ru-3’, 34% Ru-4. After 3 h: 22% Ru-3,
6% Ru-3’, 0% Ru-4; 11% missing Ru. Adding ethylene
1H NMR (C
6
D , 300 MHz, Figure S16): δ 19.53 (q, JHH =
6
3
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5.9 Hz, [Ru]=CHMe), 8.64 (br s, 4H, py o-CH), 6.78 (s, 2H,
(
1 atm) resulted in reformation of Ru-4. After ethylene
Mes CH), 6.75 (s, 2H, Mes CH), 6.71 (br s, 2H, py p-CH),
addition: 51% Ru-3, 4% Ru-3’, 29% Ru-4; 16% missing
Ru.
6.42 (br s, 4H, py m-CH), 3.50–3.37 (m, 2H, NCH
3.23 (m, 2H, NCH ), 2.78 (s, 6H, Mes CH ), 2.57 (s, 6H,
Mes CH ), 2.11 (br s, 3H, Mes CH
), 2.10 (d, 3JHH = 5.9 Hz,
3H, [Ru]=CHCH ; overlaps with Mes CH
Mes CH ).
1H NMR (CDCl
2
), 3.37–
2
3
Bimolecular Coupling of GIII. In a representative reac-
tion, solid GIII (9.6 mg, 0.013 mmol) and ca. 1 mg DMT
3
3
3
3
), 2.06 (br s, 3H,
were dissolved in 0.66 mL C
6
D6
in a J. Young NMR tube,
3
1
3
to give a final Ru concentration of 20 mM. A H NMR
spectrum was recorded to establish the initial integration
ratio of GIII relative to DMT. The NMR tube was heated
to 60 °C in a thermostatted oil bath, and decomposition
3
, 300 MHz, –20 °C): δ 19.10 (br q, JHH
=
5.7 Hz, [Ru]=CHMe), 8.70 (br d, 3JHH = 3.7 Hz, 2H, py o-
CH), 8.06 (br s, 2H, py o-CH), 7.52 (br t, 3JHH = 7.5 Hz, 1H,
py p-CH), 7.43 (br t, 3JHH = 6.6 Hz, 1H, py p-CH), 7.06 (br
s, 2H, py m-CH), 6.96 (br s, 4H, Mes CH), 6.79 (br s, 2H,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
of GIII was monitored by H NMR analysis over 5 d,
during which time the color changed from green to or-
py m-CH), 4.20–4.02 (m, 2H, NCH
NCH ), 2.55 (s, 6H, Mes CH ), 2.42 (s, 6H, Mes CH
(br s, 3H, Mes CH ), 2.21 (br s, 3H, Mes CH
= 5.7 Hz, [Ru]=CHCH , 3H).
, 300 MHz, –20 °C; Figure S18): δ
2
), 4.04–3.87 (m, 2H,
1
ange. After this time, no GIII remained. H NMR (C
6
D6
,
2
3
3
), 2.32
), 1.72 (d, 3JHH
3
1
5
00 MHz; diagnostic signals; Figures S14, S15): δ 19.66 (s,
3
3
3
H, [Ru]=CHPh of GIII; none remaining), 9.62 (dt, JHH
=
3
.2 Hz, 4JHH = 1.5 Hz, 4H, py o-CH of Ru-3; 34%), 9.16 (br
13
C{ H} NMR (CDCl
1
3
d, 3JHH = 4.2 Hz, 2H, py o-CH of Ru-3’; 28%), 9.06 (dt, 3JHH
330.1 ([Ru]=CHMe; H-detected via HSQC), 219.4 (CNHC),
150.8 (py o-CH), 149.9 (py o-CH), 139.0, 138.5, 138.1,
137.9, 136.4, 136.2 (py p-CH), 135.0 (py p-CH), 129.2 (Mes
CH), 128.9 (py m-CH), 128.2, 123.5 (Mes CH), 123.2 (py
1
=
4
7
5 Hz, 4JHH = 1.5 Hz, 8H, py o-CH of Ru-4; 1.6%), 8.00 (s,
H, CH of DMT), 7.33 (m, 4H, o-CH of (E)-stilbene; 74%),
.24 (m, 4H, o-CH of (Z)-stilbene; none present).
1
(E)-PhCH=CHPh: H NMR (C
6
D6
, 300 MHz): δ 7.33 (m,
m-CH), 51.0 (NCH
2
), 50.7 (NCH
2
), 46.4 ([Ru]=CHCH
), 19.7 (Mes CH ), 18.5
) 1484
3
),
4
H, o-CH), 7.17 (m, 4H, m-CH), 7.08 (m, 2H, p-CH), 7.01
21.1 (Mes CH
(Mes CH
(m), ν(CH
[RuCl(H
IMes)(pyrene)]•+ 645.12 (42%; calc’d: 645.16),
[Ru(H
IMes–H)(pyrene)]•+ 609.14 (67%; calc’d: 609.18),
[RuCl(H
IMes–H)]•+ 442.05 (100%; calc’d: 442.08). Sample
3
), 20.9 (Mes CH
3
3
+
-1
(s, 2H, =CH). EI MS, m/z: [M ] 180.1 (simulated: 180.1 for
3
). IR (ATR, cm ): ν(C–H) 2864 (w), ν(CH
3
1
C
14
H
12). The H NMR and GC-MS data matched those of
3
) 1405. MALDI-TOF MS (pyrene matrix), m/z:
a commercial sample (Sigma-Aldrich).
2
1
(Z)-PhCH=CHPh (Alfa Aesar, not observed): H NMR
2
(C
6
D6
, 300 MHz): δ 7.24 (m, 4H, o-CH), 7.06–6.93 (m, 6H,
H
2
m- and p-CH), 6.47 (s, 2H, =CH).
Synthesis of RuCl (H IMes)(py) (=CHMe), GIIIe. Syn-
decomposition precluded satisfactory microanalysis.
Bimolecular Coupling of GIIIe: Quantification of Bu-
tene and Ru-3/3’. Caution! Thermal expansion in these
variable-temperature experiments results in an explo-
sion hazard, because it is essential to minimize the head-
space in the NMR tube in order to measure as much as
possible of the volatile products evolved (e.g., propene,
butene). To minimize this risk, the headspace volume
required to accommodate thermal expansion of the sol-
vent and evolution of butene was calculated (see SI), and
quadrupled to provide a safety margin. Any potential
for damage to personnel or the NMR probe was limited
further by warming the sample only behind a blast
shield (not in the probe), and using a face shield when
transferring the tube to and from the NMR probe.
2
2
2
thesis of GIIIe was carried out by a modified version of
the method reported17 for the 3-bromopyridine analog.
Decomposition of GIIIe was minimized by performing
the reaction in C
6
H6 at near-freezing temperatures, and
then subliming the solvent by freeze-drying. On the
Schlenk line, a solution of GIII (100 mg, 0.138 mmol) in
C
6
H6 (2 mL) was freeze/pump/thaw degassed, then
thawed at 6 °C. Cis-2-butene (1 atm) was admitted via
the side-arm of the Schlenk flask, and the reaction was
stirred for 10 min, over which time it turned from green
to brown, and a red-brown solid precipitated. Pyridine
(100 ꢀL) was then injected, and the solution was frozen
and lyophilized. The resulting tan powder was suspend-
ed in pentane (2 mL) in the glovebox, filtered, washed
with pentane (3 x 1 mL), and dried in vacuo. Repeating
the entire process with the crude product afforded GIIIe
as a reddish-tan powder (85 mg, 93%), containing trace
Ru-3 and GIII (each ≤1%). Re-subjecting to reaction with
In a representative experiment, a solution of GIIIe (13.9
mg, 0.0209 mmol) and DMT (ca. 1 mg) in C
6
D6 (2.75 mL)
1
was added to a J. Young tube. A H NMR spectrum was
recorded to establish the initial integration ratio of GIIIe
relative to DMT. The full length of the NMR tube was
heated to 60 °C in a thermostatted water bath behind a
blast shield. A colour change from brown to orange was
observed within 25 min. After this time, the tube was
removed from the 60 °C bath, cooled in a room-
temperature water bath (1 min), and then immediately
butene, or reprecipitating from CH
2
Cl2–pentane at 0 °C,
led to competing decomposition. GIIIe is sufficiently
1
stable to observe at RT by H NMR spectroscopy, but its
spectrum degrades noticeably over the longer collection
1
3
times required for C NMR analysis (3-4 h).
1
inserted into the probe for H NMR analysis.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 18
1H NMR (C
6
D
6
, 300 MHz; diagnostic signals; Figures S18,
3
CH
2
Cl
2
(250 ꢀL) was then added dropwise. Care was
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
S20): δ 19.53 (q, JHH = 5.9 Hz, [Ru]=CHMe; none remain-
ing), 9.62 (dt, JHH = 5.2 Hz, 4JHH = 1.5 Hz, 4H, py o-CH of
Ru-3; 30%), 9.16 (br d, 3JHH = 4.2 Hz, 2H, py o-CH of Ru-
taken to dribble the solution down the cold wall of the
flask, to avoid warming the reaction mixture. A colour
change from pink to green occurred within 10 min. The
solution was stirred for an additional 20 min, then
cooled further to –78 °C (acetone–dry ice), after which
cold pentane (30 mL; –78 °C) was added via cannula. To
avoid warming, the cannula was kept cold with dry ice.
The resulting precipitate was decanted and dried in
vacuo to afford 74 mg of a pale green solid containing a
near-equimolar mixture (1.2:1.0) of Ru-6 and
3
3
’; 48%), 8.00 (s, 4H, CH of DMT), 5.52–5.30 (overlapping
m, 2H, =CH of (E)/(Z)-2-butene and 2-pentene, 53% to-
tal), 5.25 (s, 4H, C , <1%), 5.06–4.90 (m, 2H, =CH of
2
H4
2
propene, 23%). The overlapping olefinic signals for 2-
butene and 2-pentene were deconvoluted by re-running
the spectrum at 600 MHz (Figure S19). The observed
proportion of butene, pentene, and propene increased on
warming the sample to 60 °C in the probe, perhaps re-
flecting pressure buildup in the headspace on warming.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
[H
2
C=CHPCy
3
]OTf,10a with ca. 10-15% unidentified Ru-
H
2
IMes impurities. Yield of crude Ru-6: 75%. Attempts
1H NMR (C
6
D
6
, 300 MHz, 60 °C; diagnostic signals): δ
.52–5.30 (m, 2H, =CH of (E)/(Z)-butene and 2-pentene,
9%), 5.06–4.88 (m, 2H, =CH of propene, 24%). Adding 2
at reprecipitation or washing caused decomposition into
Ru-7. Given below are NMR details for Ru-6 alone; those
for the phosphonium salt appear in the supporting in-
formation.
5
6
2
equiv pyridine to the reaction resulted in conversion of
the Ru products to Ru-3 (85%). The identities of the bu-
tene, pentene, and propene products were confirmed by
1H NMR (CDCl
(s, 2H, [Ru]=CH
13H, Mes CH and o-dianiline CH; overlaps with residual
CHCl
), 6.57 (br d, 3JHH = 7.8 Hz, 1H, o-dianiline CH), 6.44
(br d, 3JHH = 7.8 Hz, 1H, o-dianiline CH), 4.37 (br s, 2H,
NH ), 4.14–3.70 (m, 4H, NCH ), 3.68 (br s, 1H, NH ),
3.41 (br s, 1H, NH ), 2.67 (s, 3H, CH ), 2.50 (s, 3H,
), 2.35 and 2.34 (overlapping s, 9H, CH ), 2.23 (s, 3H,
). Assignment of the NH signals was confirmed by
O at –50 °C, and rapidly shaking the
tube (1-2 sec) before re-immersing in the cold-bath.
3
, 300 MHz, –50 °C; Figure S22): δ 19.46
), 7.39 (s, 1H, Mes CH), 7.32–6.92 (m,
2
1H and 13C{ H} NMR, H– H COSY, H– C HSQC, and
1
1
1
1
13
1
13
H– C HMBC analysis. The key NMR shifts are provid-
3
ed below.
1
(E)-2-butene: H NMR (C
6
D
6
, 600 MHz): δ 5.38 (m, 2H,
2
2
a
H
b
=
CH), 1.57 (m, 6H, CH
3
). 13C{ H} NMR (C
1
6
D
6
, 150 MHz):
, 600 MHz): δ
).
a
Hb
3
1
δ 124.5, 17.8. (Z)-2-butene: H NMR (C
5.48 (m, 2H, =CH), 1.51 (d, JHH = 5.0 Hz, 6H, CH
6
D
6
CH
3
3
3
3
3
CH
2
13
1
C{ H} NMR (C
signals are slightly shifted relative to the reported values
in CDCl
,50 but are otherwise in good agreement.
6
D6
, 150 MHz): δ 125.7, 12.0. The NMR
adding ca. 5 ꢀL D
2
13
1
3
C{ H} NMR (CDCl
3
, 77.5 MHz, –50 °C; Figure S23): δ
1
1
Propene: H NMR (C
6
D
6
, 600 MHz): δ 5.76–5.67 (m, 1H,
314.2 ([Ru]=CH
2
, H-detected by HMQC), 218.5 (CNHC),
CH), 5.01 (dm, 3JHH = 17.0 Hz, =CH
2
), 4.95 (dm, JHH
), 1.55 (dt, 3JHH = 6.5 Hz, 4JHH = 1.5 Hz).
, 150 MHz): δ 134.0 (=CH), 115.6
). The observed shifts are in excellent
3
=
143.6, 139.8, 139.0, 138.9, 138.6, 138.4, 137.1, 134.8, 134.3,
131.1, 130.6, 130.4, 129.4, 129.2, 128.9, 128.4, 119.1, 115.8,
=
1
0.3 Hz, =CH
2
13
1
C{ H} NMR (C
6
D6
50.6 (NCH
(CH ), 19.0 (CH
Synthesis of RuCl
2
), 49.8 (NCH
), 18.6 (CH
(H IMes)(py)
2
), 21.5 (CH
), 18.5 (CH
(=CH ) (GIIIm: n = 2;
3
), 21.4 (CH
3
), 19.2
(=CH
2
), 19.4 (CH
3
3
3
3
3
).
agreement with the reported values.49
2
2
n
2
1
(E)-2-pentene: H NMR (C
6
D6
, 600 MHz): δ 5.52–5.33 (m,
GIIIm’: n = 1). The synthesis was undertaken using the
method for Ru-6 above, with several modifications to
prevent decomposition and enable precipitation of the
product: (i) use of lower temperatures (down to –116
°C); (ii) immediate workup after adding pyridine; and
(iii) higher Ru concentrations (ca. 110 mM starting PII),
to aid in precipitating the product. Use of stoichiometric
pyridine was likewise essential, to prevent oiling out of
the product.
2
=
1
H, =CH), 1.94 (dq, 3JHH = 7.5 Hz, 3JHH = 1.3 Hz, 2H,
). 13C{ H} NMR (C
). The NMR signals are slight-
, but
1
6
D6, 150 MHz): δ 133.0 (=CH),
CHCH
2
22.8 (=CH), 26.0 (=CHCH
2
ly shifted relative to the reported values in CDCl
3
5
1
are otherwise in good agreement. A predominantly (E)-
configuration is assigned on the basis of the downfield
location of the =CHCH
cf. 20.0 for (Z)-pentene in CDCl
Synthesis of RuCl (H
2
singlet (δ
C
26.0, vs. 25.2 in CDCl ;
3
3
).51
2
2
IMes)(o-dianiline)(=CH
2
), Ru-6.
In a 10 mL Schlenk flask, a 110 mM solution of PII (60
For the NMR-scale synthesis of Ru-6, see SI. In a 10 mL
Schlenk flask, a 60 mM solution of PII (75 mg, 0.081
mg, 0.065 mmol) in CH
2
Cl2 (0.60 mL) was prepared and
frozen in N (l). The headspace was evacuated, the flask
2
mmol) in CH
2
Cl
2
(1.10 mL) was prepared and frozen in
was sealed, and the solution was allowed to thaw at –50
°C. Ethylene gas (1 atm) was then admitted via the side-
arm, after which the flask was sealed again, and the
reaction was stirred at –50 °C for 10 min. A colour
change from dark brown to bright pink occurred over
this time. The solution was cooled further, to –78 °C, and
a solution of pyridine (10.3 ꢀL, 0.132 mmol, 2.0 equiv) in
pentane (150 ꢀL) was added dropwise. A colour change
N
2
(l). The headspace was evacuated, the flask was
sealed, and the solution was allowed to thaw at –50 °C.
Ethylene gas (1 atm) was then admitted via the side-arm,
after which the flask was sealed again, and the reaction
was stirred at –50 °C for 10 min, over which time the
colour changed from dark brown to bright pink. A solu-
tion of o-dianiline (15.8 mg, 0.0855 mmol, 1.05 equiv) in
ACS Paragon Plus Environment

Page 13 of 18
Journal of the American Chemical Society
from pink to green occurred over the course of the addi-
tion. Cold pentane (8 mL; –116 °C; N (l)–ethanol bath)
°C), and inserted into an NMR probe pre-cooled to –20
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
°C, to measure the initial integration ratio of Ru-6 vs.
DMT. The sample was then ejected from the NMR
probe, mixed briefly by shaking, and allowed to warm in
a 23 °C water bath (blast shield; see above). Decomposi-
was then added via a cannula chilled with dry ice. The
resulting precipitate was decanted and dried in vacuo to
afford 59 mg of the mustard-yellow product accompa-
1
nied by [H
unidentified Ru-H
additional mesityl CH
2
C=CHPCy
IMes impurities, which give rise to
peaks (Figure S27). Integration of
3
]Cl, in a 2:3 ratio. Also present are
tion of Ru-6 was monitored over 1 h at RT ( H NMR).
2
During this time the colour changed from green to
1
3
brown. After 6 min: H NMR (CD
2
Cl2, 300 MHz, RT;
the alkylidene singlets at –50 °C indicates a ca. 1:2 ratio
of GIIIm and GIIIm’. Crude yield: ca. 60% based on
GIIIm/m’. Attempts to purify by washing with cold
diagnostic signals only): δ 19.53 (s, 2H, [Ru]=CH
6; 27%), 8.09 (s, 4H, CH of DMT), 5.40 (s, 4H, C
4.91 (d, 2JHH = 10.3 Hz, 1H, NH
Ru-6, 52% C , 61% Ru-7. At 85 min: no Ru-6 remain-
ing, 53% C , 55% Ru-7.
RuCl (H IMes)(o-dianiline) Ru-7. H NMR (CD
2
of Ru-
2
H4
; 63%),
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
of Ru-7; 55%). At 1 h: 2%
1
pentane (–116 °C) afforded no improvement. H NMR
2
H
4
analysis was carried out at 0 °C in CDCl
3
(in which the
2
H
4
1
complex is more stable than in CD Cl ). Exchange aver-
2
2
2
2
2
Cl2, 300
aging at this temperature results in a single alkylidene
peak, greatly simplifying analysis. Sample decomposi-
MHz; Figure S25): δ 7.29–6.95 (m, 9H, Ar CH of o-
dianiline and Mes CH), 6.94 (s, 2H, Mes CH), 6.86 (d, 1H,
3JHH = 7.7 Hz, Ar CH of o-dianiline), 6.18 (m, 1H, Ar CH
1
3
1
tion occurs over 3–4 h, precluding C{ H} NMR analysis.
1H NMR (CDCl
, 300 MHz, 0 °C; Figure S27): δ 18.83 (s,
H, [Ru]=CH ), 8.67 (br s, 0.6H, py o-CH), 7.82 (br s, 2H,
py o-CH), 7.51 (br s, 0.6H, py m-CH), 7.37 (br t, 0.3H, 3JHH
3
of o-dianiline), 4.89 (d, 2JHH = 10.4 Hz, 1H, NH
(d, 2JHH = 10.4 Hz, 1H, NH
), 4.12–3.89 (m, 6H, NCH
and NH ), 2.56 (br s, 6H, Mes o-CH ), 2.41 (s, 6H, Mes
o-CH ), 2.25 (s, 6H, Mes p-CH ). The two NH doublets
that overlap with the backbone NCH protons of H IMes
a
Hb), 4.20
2
2
aH
b
2
a
H
b
3
=
7
4
7.0 Hz, py p-CH), 7.15 (br t, 3JHH = 7.0 Hz, 1H, py p-CH),
3
3
.01 (s, 4H, Mes CH and py m-CH), 6.95 (s, 2H, Mes CH),
2
2
1
1
.12 (m, 2H, NCH
2
), 4.02 (m, 2H, NCH
2
), 2.55 (s, 6H,
were located from their H– H COSY correlations with
CH
3
), 2.35 (s, 6H, CH
3
), 2.32 (s, 6H, CH ). Exchange aver-
3
the well-separated NH signals further downfield. The
aging of the bis- and mono-pyridine complexes results in
a single set of signals at 0 °C (as also seen for GIII and its
mono-pyridine derivative GIII’; see below). At –50 °C,
pyridine exchange is retarded and two sets of signals
diastereotopic NH
a
H
b
pairs thus identified appear at:
4.89 / 3.95 ppm, and 4.20 / 4.04 ppm. 13C{ H} NMR
1
(CH
2
Cl2, 77.5 MHz; Figure S26): δ 213.8 (CNHC), 140.0,
139.7, 138.9 (br), 138.3 (br), 138.0, 130.7, 129.8 (br), 129.8
1
emerge. H NMR (CDCl
9.35 (br s, [Ru]=CH for GIIIm, 30%), 18.63 (br s,
Ru]=CH for GIIIm’, 70%). Adding pyridine (10 ꢀL) to
the NMR sample at –50 °C caused clean conversion to
3
, –50 °C; key signals only): δ
(Mes CH), 129.7, 129.4, 128.1, 127.9, 127.54, 127.45, 124.8,
1
2
124.2, 123.7, 122.9, 122.6, 122.3, 52.3 (NCH
2
), 20.6 (Mes p-
). MALDI-
[
2
CH ), 18.6 (br, Mes o-CH ), 18.3 (Mes o-CH
3
3
3
+
TOF MS (pyrene matrix), m/z: [Ru-7–H] 661.141 (calc’d:
661.144).
GIIIm. Key 13C NMR signals (detected via H– C
HMQC in CDCl at 0 °C): 305.1 (s, [Ru]=CH ), 129.7 (s,
Mes CH), 129.4 (s, Mes CH), 51.1 (s, NCH ), 49.9 (s,
NCH ).
1
13
3
2
Bimolecular coupling of GIIIm: quantification of eth-
ylene (Figure S28). Experiment carried out as for Ru-6
above, using solid GIIIm/m’ (ca. 60 mg, 0.038 mmol) and
2
2
Bimolecular coupling of Ru-6: quantification of eth-
ylene and formation of Ru-7. These experiments were
carried out with the precautions against explosion de-
scribed above. The sample was prepared in the glove-
box, so that the septum-sealed, screw-capped NMR tube
could be replaced with a J. Young NMR tube (see Figure
S1; 2.15 mL capacity), and perturbation by oxygen could
be more rigorously inhibited. While the lower operating
temperature is then limited to –35 °C (the temperature of
the glovebox freezer), this was deemed acceptable given
rapid manipulation, as decomposition commences only
at ca. –10 °C. In a representative experiment, crude Ru-6
TMB (0.47 mg) in CD
2
Cl2 (1.72 mL). After 20 min: δ 18.80
(s, 2H, [Ru]=CH of GIIIm; none remaining), 8.99 (br d,
2
2JHH = 5.6 Hz, 2H, py o-CH of Ru-3, 12%), 8.60 (br s, 2H,
py o-CH of Ru-3’, 4%), 6.07 (s, 3H, Ar CH of TMB), 5.40
(s, 4H, C
2
H4; 70%).
Kinetics studies. The rate of disappearance of the me-
thylidene signal for Ru-6 vs. DMT internal standard was
measured as in the ethylene quantitation experiments
above, but using a standard NMR-tube headspace, as
quantification of evolved gases is irrelevant.
Kinetics at High [Ru]
was chilled in a sand-bath in the glovebox, loaded with
solid Ru-6 (17 mg, 0.012 mmol), cold CD Cl (600 μL),
0
(ca. 20 mM Ru-6). A J. Young tube
(
ca. 30 mg, 0.022 mmol) and 1.7 mg DMT (40 ꢀL of a 0.22
M stock solution in CD Cl ) were added to a cold NMR
tube bedded in a chilled (–35 °C) sand-bath. Cold CD Cl
1.72 mL; used instead of C to improve the solubility
2
2
2
2
and DMT (ca. 1 mg) in the glovebox, to give a Ru con-
2
2
centration of approximately 20 mM (based on Ru-6 is
1
(
7
D8
48% pure, based on H NMR analysis indicating 1.15
of Ru-6) was added, and the solution was mixed using a
chilled pipette. The tube was sealed, transferred to the
instrument room in an ethylene glycol–dry ice bath (–20
equiv [H
2
C=CHPCy
3
]OTf and ca. 15% Ru-H IMes impu-
2
rities in this particular sample). Decomposition of Ru-6
was monitored over three half-lives. A linear depend-
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 18
ence on [Ru-6]-1 (Figures 6 and S9) indicates that cou-
pling is second-order at this concentration (kobs = 0.12 ±
confirming either a single imaginary frequency (for the
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
transition state) or no imaginary frequencies (for mini-
ma). The only exception was Ru-11, for which the pro-
cedure described above returned an imaginary frequen-
-1
-1
0
.02 M s , based on two trials).
Kinetics at Low [Ru] (ca. 1 mM Ru-6). As above, but with
ca. 1 mg solid Ru-6 and DMT (0.024 mg, 0.125 nmol; 5.0
L of a 26 mM stock solution in CD Cl ) in 600 ꢀL
CD Cl . A starting Ru concentration of ca. 1 mM was
0
-1
cy i13cm . This imaginary frequency was confirmed to
be an artifact resulting from the above-mentioned de-
fault reduction of the grid quality in analytical frequency
calculations in Gaussian 09.62 Specifically requesting the
“ultrafine” grid also in the CPHF-based frequency calcu-
lation confirmed all-positive curvature for this interme-
diate as well. The thermal correction for Ru-11 was thus
obtained from the latter vibrational analysis using the
“ultrafine” integration grid, while the standard proce-
dure described above was used for all species for which
artifacts from the grid were not detected. The transla-
tional, rotational, and vibrational components of the
thermal corrections to enthalpies and Gibbs free energies
were calculated within the ideal-gas, rigid-rotor, and
harmonic oscillator approximations, except that all fre-
ꢀ
2
2
2
2
established via integration of the methylidene singlet vs.
DMT at –20 °C. Decomposition of Ru-6 was monitored
at +10 °C to permit collection of sufficient scans for good
signal-to-noise ratios. Analysis as above revealed a linear
dependence on ln[Ru-6], indicating that bimolecular
coupling is first-order in Ru-6 at low concentrations (kobs
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-4 -1
=
(1.9 ± 0.3)*10 s , based on two trials).
Rate Inhibition by Added o-Dianiline at Low [Ru]
0
. Experi-
ments with and without exogenous o-dianiline were
conducted at RT, under conditions that otherwise corre-
spond to the 1 mM experiments above. The proportion
of Ru-6 was assessed after 1 h. Control: 5% Ru-6 remain-
ing. With 4 equiv o-dianiline: 21% Ru-6.
-1
−1
quencies below 100 cm were shifted to 100 cm when
calculating the vibrational component of the entropy,63,64
to prevent asymptotic behavior of the harmonic approx-
imation for modes of very low frequencies (an approach
termed the “quasi-harmonic approximation” in the fol-
lowing).
Computational methods. Geometry optimization. All DFT
calculations were performed with the Gaussian 09 suite
of programs.52 Geometry optimization were performed
using Head-Gordon's long-range- and dispersion-
corrected hybrid density functional ωB97XD53-55 as im-
plemented in Gaussian 09. This functional was chosen as
it provides geometries in very good agreement with
those of X-ray diffraction analysis of Ru metathesis cata-
Intrinsic reaction coordinate (IRC)65 calculations (using
the local quadratic approximation algorithm (LQA)66
connected intermediate Ru-8 with transition state Ru-9.
Single-point energy calculations. All single-point energy
calculations were performed with the Gaussian 09 im-
plementation of the generalized gradient approximation
functional of Perdew, Burke and Ernzerhof (PBE),67 and
included Grimme’s D3 empirical dispersion term68 with
revised Becke-Johnson69 damping parameters (together
labeled PBE-D3M(BJ), for brevity).70 Ruthenium was
described by the ECP28MDF relativistic effective core
potential,57 accompanied by a correlation-consistent
valence quadruple-ζ plus polarization basis set
56
lysts and other homogeneous catalysts. Ruthenium was
described by the Stuttgart 28-electron relativistic effec-
tive core potential, termed ECP28MDF and retrieved
from the Stuttgart/Cologne group website,57,58 in combi-
nation with the accompanying correlation-consistent
valence double-ζ plus polarization basis set (cc-pVDZ-
PP)57 retrieved from the EMSL basis set exchange data-
base.59 All remaining atoms were described by Dun-
ning's correlation-consistent valence double-ζ plus po-
larization basis sets (cc-pVDZ),60,61 as retrieved from the
EMSL basis set exchange database.59 Numerical integra-
tion was performed using the “ultrafine” grid of Gaussi-
an 09. This grid specification defaults to the coarser
“SG1” grid for analytical Hessian calculations using the
CPHF procedure. The built-in Gaussian 09 stability
check was carried out for all self-consistent field solu-
tions prior to geometry optimization. Instable solutions
were re-optimized to real, spin-restricted solutions. Ge-
ometries were then optimized using tight convergence
criteria (max. force 1.5·10−5 a.u., RMS force 1.0·10−5 a.u.,
max. force 6.0·10−5 a.u., RMS force 4.0·10−5 a.u.), without
symmetry constraints, and using default convergence
criteria for the self-consistent field (SCF) optimization
(ECP28MDF_VQZ),57
both
obtained
from
the
Stuttgart/Cologne group website. Carbon and hydrogen
atoms were described by valence quadruple-ζ plus po-
59
60
larization (EMSL: cc-pVQZ) basis sets. All other atoms
were described by the valence quadruple-ζ plus polari-
zation augmented with diffuse functions (EMSL: aug-cc-
pVQZ),59,60,71 Electrostatic and non-electrostatic solvation
effects in dichloromethane were taken into account us-
ing the polarizable continuum model (PCM), in combi-
nation with the “Dis”, “Rep”, and “Cav” keywords and
the built-in program values (dielectric constant, number
density, etc.).72 The solute cavity was constructed using
the united atom topological model with atomic radii
optimized for Hartree−Fock (termed “UAHF”)72d,73 Nu-
merical integrations were performed with the “ultrafine”
grid of Gaussian 09 and the self-consistent field (SCF)
−8
procedure (RMS change in density matrix < 1.0·10 , max.
−
6
change in density matrix = 1.0·10 ).
All stationary points were characterized by the eigen-
values of the analytically calculated Hessian matrix,
-5
density-based convergence criterion was set to 10 (RMS
ACS Paragon Plus Environment

Page 15 of 18
Journal of the American Chemical Society
−
5
change in density matrix < 1.0·10 , max. change in densi-
This work was funded by NSERC of Canada and by the
Research Council of Norway (RCN, project 262370). NSERC
is thanked for fellowships to GAB and CSH, and RCN for a
fellowship to MF, as well as CPU (NN2506K) and storage
resources (NS2506K). Giovanni Occhipinti is thanked for
useful discussions.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
−
3
ty matrix = 1.0·10 ).
Calculation of Gibbs free energies. Gibbs free energies were
calculated at 298.15
K
according to equation 1.
CH Cl
CH Cl
T=298.15K
T=298.15K
2
PBE-D3M(BJ)
2
2
PBE-D3M(BJ)
2
(1)
G
=E
+ꢁGϖ +ꢁG1atm→1M
B97XD,qh
CH Cl
Here
E
2
PBE-D3M(BJ)
2
corresponds to the potential energies
REFERENCES
resulting from single point calculations with PBE-
(
1)
(a) Grela, K., Olefin Metathesis-Theory and Practice.
D3M(BJ), including the contributions from the implicit
Wiley: Hoboken, NJ, 2014. (b) Grubbs, R. H.; Wenzel, A. G.,
T=298.15K
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
solvation model; ꢁGϖ
is the thermal correction to
B97XD,qh
Handbook of Metathesis. 2nd ed.; Wiley-VCH: Weinheim, 2015.
(2)
(a) Higman, C. S.; Lummiss, J. A. M.; Fogg, D. E.
the Gibbs free energy, calculated at the geometry opti-
mization level with the quasi-harmonic approximation,
Angew. Chem., Int. Ed. 2016, 55, 3552–3565. (b) K. R. Fandrick, J.
Savoie, N. Y. Jinhua, J. J. Song, C. H. Senanayake, In ref 1a, pp.
349-366. (c) Farina, V.; Horváth, A. In ref 1a, pp 633–658.
T=298.15K
and ꢁG1atm→1M is the standard-state correction corre-
-3
-1
sponding to 1 M solution (but exhibiting infinite-
dilution, ideal-gas-like behavior). The latter is equal to
(3)
Rate constants for initiation (k ; x 10 s ) at 5 °C: GII,
i
0.0032; HII, 2.6; GIII, >200. See: Love, J. A.; Morgan, J. P.;
Trnka, T. M.; Grubbs, R. H. Angew. Chem., Int. Ed. 2002, 41,
.89 kcal mol−1 (= RT·ln(24.46)). Table S5 reports all these
1
4
035–4037, and references therein. % Recapture = proportion of
the resting-state species at 50% catalyst decomposition during
styrene metathesis (100 equiv styrene, 22 °C, C ). Data for
selected Class A–C catalysts: GII, 15%; HII, 100%; GIII, 0%.
The resting states are, respectively, RuCl (H IMes)(PCy )(=CH
GIIm, HII itself, and (see below) RuCl (H IMes)(py) (=CH ).
(4) Higman, C. S.; Nascimento, D.; Ireland, B. J.;
Audorsch, S.; Bailey, G. A.; Fogg, D. E. J. Am. Chem. Soc. 2018,
40, 1604−1607.
5) For evidence of “boomerang” recapture of
values, together with the single-point energy calculated
at the geometry optimization level ( EϖB97XD ) and the
corresponding Gibbs free energy including thermal cor-
6
D
6
2
2
3
2)
T=298.15K
rections from the harmonic approximation ( GϖB97XD ).
2
2
n
2
Natural bond orbital analysis. The natural bond orbital
analyses were performed using the NBO 6.0 program.74
Memory restrictions prohibited NBO calculations on the
dimeric ruthenium complexes (Ru-8, Ru-13 and transi-
tion state Ru-9) using the very large basis sets for single-
point calculations described above. Instead, Dunning's
correlation-consistent valence triple-ζ plus polarization
basis sets (cc-pVTZ-DK)57,60,61 were used for all atoms in
the single-point, self-consistent field calculations giving
the electron density used in the NBO analyses. The re-
maining settings of the DFT model and the self-
consistent field protocol were as described above for the
single-point energy calculations, the sole exception being
use of the original Becke-Johnson69 damping parameters
when calculating Grimme’s D3 empirical dispersion
terms.68
1
(
isopropoxystyrene, see: (a) Bates, J. M.; Lummiss, J. A. M.;
Bailey, G. A.; Fogg, D. E. ACS Catal. 2014, 4, 2387−2394. (b)
Griffiths, J. R.; Keister, J. B.; Diver, S. T. J. Am. Chem. Soc. 2016,
1
38, 5380−5391. For the original proposal, see: (c) Garber, S. B.;
Kingsbury, J. S.; Gray, B. L.; Hoveyda, A. H. J. Am. Chem. Soc.
000, 122, 8168–8179.
(6) The vinylphosphonium salt [H C=CHPCy ]Cl lost on
2
2
3
initiation of PII does not re-enter metathesis. See: Leitao, E. M.;
van der Eide, E. F.; Romero, P. E.; Piers, W. E.; McDonald, R. J.
Am. Chem. Soc. 2010, 132, 2784–2794.
(7)
(a) Schrodi, Y., Mechanisms of Olefin Metathesis
Catalyst Decomposition and Methods of Catalyst Reactivation.
In ref 1b, pp 323–342. (b) Chadwick, J. C.; Duchateau, R.; Freixa,
Z.; van Leeuwen, P. W. N. M., Alkene Metathesis. In
Homogeneous Catalysts: Activity – Stability – Deactivation, Wiley-
VCH: Weinheim, 2011; pp 347-396.
ASSOCIATED CONTENT
Supporting Information
(8)
(a) Hong, S. H.; Wenzel, A. G.; Salguero, T. T.; Day,
M. W.; Grubbs, R. H. J. Am. Chem. Soc. 2007, 129, 7961–7968. (b)
Hong, S. H.; Day, M. W.; Grubbs, R. H. J. Am. Chem. Soc. 2004,
Experimental procedures and NMR spectra, sample DFT
input files, calculated energies, molecular structures and
other properties (PDF). A file containing 3D rotatable imag-
es of all geometry-optimized structures as well as the IRC
trajectory connecting transition state Ru-9 with intermedi-
ate Ru-8 (XYZ). This material is available free of charge via
the Internet at http://pubs.acs.org.
1
26, 7414–7415.
9) (a) McClennan, W. L.; Rufh, S.; Lummiss, J. A. M.;
(
Fogg, D. E. J. Am. Chem. Soc. 2016, 138, 14668−14677. (b)
Lummiss, J. A. M.; McClennan, W. L.; McDonald, R.; Fogg, D.
E. Organometallics 2014, 33, 6738–6741. (c) Lummiss, J. A. M.;
Ireland, B. J.; Sommers, J. M.; Fogg, D. E. ChemCatChem 2014, 6,
4
59–463. (d) Lummiss, J. A. M.; Botti, A. G. G.; Fogg, D. E. Catal.
AUTHOR INFORMATION
Corresponding Author
Sci. Technol. 2014, 4, 4210–4218.
(10) (a) Romero, P. E.; Piers, W. E. J. Am. Chem. Soc. 2005,
1
27, 5032–5033. (b) Romero, P. E.; Piers, W. E. J. Am. Chem. Soc.
*
Email: dfogg@uottawa.ca.
2007, 129, 1698–1704. (c) van der Eide, E. F.; Piers, W. E. Nature
Chem. 2010, 2, 571–576.
ACKNOWLEDGEMENT
(11) Solans-Monfort, X.; Coperet, C.; Eisenstein, O. J. Am.
Chem. Soc. 2010, 132, 7750–7757.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 18
(
12) Nizovtsev, A. V.; Afanasiev, V. V.; Shutko, E. V.;
cyclohexene revealed no norcarane resulting from
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
cyclopropanation by 1H NMR or GC-MS analysis. Instead,
ethylene is generated, consistent with decomposition by
bimolecular coupling. A further possibility suggested by a
referee, involving loss of alkylidene as the free carbene on
binding a π-acceptor ligand, was explicitly ruled out in the
Cavallo study (ref 26a) on the basis of the high energy barriers
involved. Diver’s studies likewise suggested that alkylidene
insertion is intramolecular (see ref 25b).
Bespalova, N. B. NATO Sci. Ser. II 2007, 243, 125–135.
(13) Propene loss was also reported on reaction of GII
with ethylene. See: van Rensburg, W. J.; Steynberg, P. J.; Meyer,
W. H.; Kirk, M. M.; Forman, G. S. J. Am. Chem. Soc. 2004, 126,
14332–14333. Methylidene abstraction (refs 8,9) was apparently
prevented by oxidation of PCy
3
; O=PCy
3
was detected.
(
14) Bailey, G. A.; Lummiss, J. A. M.; Foscato, M.;
Occhipinti, G.; McDonald, R.; Jensen, V. R.; Fogg, D. E. J. Am.
Chem. Soc. 2017, 139, 16446–16449.
(28) Sanford, M. S.; Love, J. A.; Grubbs, R. H.
Organometallics 2001, 20, 5314–5318.
(
15) Ireland, B. J.; Dobigny, B. T.; Fogg, D. E. ACS Catal.
015, 5, 4690−4698.
16) Engel, J.; Smit, W.; Foscato, M.; Occhipinti, G.;
Törnroos, K. W.; Jensen, V. R. J. Am. Chem. Soc. 2017, 139,
6609–16619.
17) Williams, J. E.; Harner, M. J.; Sponsler, M. B.
Organometallics 2005, 24, 2013–2015.
18) Pyridine is significantly less nucleophilic than either
PPh or PCy . N values on the logarithmic Mayr nucleophilicity
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
(29) Walsh, D. J.; Lau, S. H.; Hyatt, M. G.; Guironnet, D. J.
Am. Chem. Soc. 2017, 139, 13644–13647.
(
(30) Adding pyridine to the solution, with the intention of
converting all Ru species present to Ru-3, left a shortfall of ca.
25%. Ru nanoparticles may account for the balance of material:
see (a) Higman, C. S.; Lanterna, A. E.; Marin, M. L.; Scaiano, J.
C.; Fogg, D. E. ChemCatChem 2016, 8, 2446–2449. The chief
alternative possibility, formation of paramagnetic Ru
byproducts, is improbable given the absence of significant
signal broadening in the NMR spectrum (Figure S14).
(31) Lummiss, J. A. M.; Beach, N. J.; Smith, J. C.; Fogg, D.
E. Catal. Sci. Technol. 2012, 2, 1630–1632.
1
(
(
3
3
scale: pyridine, 12.90; PPh
dichloromethane. See:
3
,
14.33; PCy , 14.64; all in
3
http://www.cup.lmu.de/oc/mayr/reaktionsdatenbank/.
19) (a) Schrock, R. R.; Copéret, C. Organometallics 2017, 36,
884–1892. An important early analysis noted the striking
propensity of M=CH species to eliminate ethylene: (b)
(
1
(32) (a) Schwab, P.; Grubbs, R. H.; Ziller, J. W. J. Am. Chem.
Soc. 1996, 118, 100–110. (b) Werner, H.; Stuer, W.;
Weberndorfer, B.; Wolf, J. Eur. J. Inorg. Chem. 1999, 1707–1713.
(c) Burrell, A. K.; Clark, G. R.; Rickard, C. E. F.; Roper, W. R.;
Wright, A. H. J. Chem. Soc., Dalton Trans. 1991, 609–614.
2
Merrifield, J. H.; Lin, G.-Y.; Kiel, W. A.; Gladysz, J. A. J. Am.
Chem. Soc. 1983, 105, 5811–5819. For crystallographically
characterized examples of methylidene complexes derived
from group 6 metathesis catalysts, see: (c) Arndt, S.; Schrock, R.
R.; Muller, P. Organometallics 2007, 26, 1279–1290. (d) Tsang, W.
C. P.; Jamieson, J. Y.; Aeilts, S. L.; Hultzsch, K. C.; Schrock, R.
R.; Hoveyda, A. H. Organometallics 2004, 23, 1997–2007.
(33) PCy
3
loss from GIIm is >40,000x slower than from GI,
RuCl (PCy (=CHPh). See: (a) Lummiss, J. A.; Higman, C. S.;
2
3)2
Fyson, D. L.; McDonald, R.; Fogg, D. E. Chem. Sci. 2015, 6, 6739–
6746. (b) Lummiss, J. A.; Perras, F. A.; Bryce, D. L.; Fogg, D. E.
Organometallics 2016, 35, 691–698.
(
20) (a) Fürstner, A.; Ackermann, L.; Gabor, B.; Goddard,
R.; Lehmann, C. W.; Mynott, R.; Stelzer, F.; Thiel, O. R. Chem. –
Eur. J. 2001, 7, 3236–3253. (b) Huang, J.; Schanz, H.-J.; Stevens,
E. D.; Nolan, S. P. Organometallics 1999, 18, 5375–5380.
(34)
The excellent Lewis basicity of o-dianiline is evident from
over a dozen reports of its Ru complexes. Selected examples: (a)
Crimmin, M. R.; Bergman, R. G.; Toste, F. D. Angew. Chem., Int. Ed.
2
2
011, 50, 4484–4487. (b) Faller, J. W.; Fontaine, P. P. Organometallics
005, 24, 4132–4138. (c) Aikawa, K.; Mikami, K. Angew. Chem., Int.
(
21) Adlhart, C.; Chen, P. Helv. Chim. Acta 2003, 86, 941–
49.
22) Experiments in 2.00 mL total volume of C D plus
9
Ed. 2003, 42, 5455–5458. A pK of 3.81 in 70% aqueous EtOH is
a
(
6
6
reported for the conjugate acid of o-dianiline; cf. 3.66 for pyridine.
See, respectively: (a) Grantham, P. H.; Weisburger, E. K.;
Weisburger, J. H. J. Org. Chem. 1961, 26, 1008–1017. (b) McDaniel,
styrene, in a 2.15 mL J. Young tube. Further reductions in
headspace present an explosion hazard (see SI).
(
23) (a) Borguet, Y.; Sauvage, X.; Zaragoza, G.;
D. H.; Özcan, M. J. Org. Chem. 1968, 33, 1922–1923. The pK
a
Demonceau, A.; Delaude, L. Organometallics 2011, 30, 2730–2738
and references therein. (b) Wolf, J.; Thommes, K.; Briel, O.;
Scopelliti, R.; Severin, K. Organometallics 2008, 27, 4464–4474. (c)
Quebatte, L.; Solari, E.; Scopelliti, R.; Severin, K. Organometallics
reported for pyridine in acetonitrile is considerably higher: see ref
24.
(
35) Nickel, A.; Ung, T.; Mkrtumyan, G.; Uy, J.; Lee, C. W.;
Stoianova, D.; Papazian, J.; Wei, W.-H.; Mallari, A.; Schrodi, Y.;
2
005, 24, 1404–1406.
(24) Reported pK
18.5. See: Cox, B. G., Acids and Bases: Solvent Effects on Acid-Base
Strength. Oxford University Press: Croydon, 2013, p. 99–115
(25) (a) Griffiths, J. R.; Hofman, E. J.; Keister, J. B.; Diver, S. T.
Pederson, R. L. Top. Catal. 2012, 55, 518–523.
a
values in acetonitrile. Pyridine: 12.6; NEt :
3
(
36) Schweitzer, D.; Snell, K. D. Org. Process Res. Dev. 2015,
9, 715−720.
37) Vignon, P.; Vancompernolle, T.; Couturier, J.-L.;
Dubois, J.-L.; Mortreux, A.; Gauvin, R. M. ChemSusChem 2015,
, 1143–1146.
38) For PCy
1
.
(
Organometallics 2017, 36, 3043–3052. (b) Galan, B. R.; Pitak, M.;
Gembicky, M.; Keister, J. B.; Diver, S. T. J. Am. Chem. Soc. 2009, 131,
8
(
3
-stabilized catalysts, the greater steric bulk of
6
822–6832. (c) Galan, B. R.; Gembicky, M.; Dominiak, P. M.; Keister,
J. B.; Diver, S. T. J. Am. Chem. Soc. 2005, 127, 15702–15703.
26) For computational studies of Buchner insertion of
the ethylidene ligand, vs. methylidene, also retards abstraction
of this ligand by phosphine. See ref 9a.
(
(
39) MacNaughtan, M. L.; Gary, J. B.; Gerlach, D. L.;
methylidene: (a) Poater, A.; Ragone, F.; Correa, A.; Cavallo, L. J.
Am. Chem. Soc. 2009, 131, 9000-9006. Of benzylidene: (b) Poater,
A.; Cavallo, L. Theor. Chem. Acc. 2012, 131, 1155.
Johnson, M. J. A.; Kampf, J. W. Organometallics 2009, 28, 2880–
2887.
(40) Leitao, E. M.; Dubberley, S. R.; Piers, W. E.; Wu, Q.;
McDonald, R. Chem. – Eur. J. 2008, 14, 11565–11572.
(
27) Primary evidence is the unperturbed H
2
IMes ligand
in Ru-3 (quantitative). In addition, reaction of GIIIm/m’ with
ACS Paragon Plus Environment

Page 17 of 18
Journal of the American Chemical Society
(
41) For selected examples of transoid diruthenium
(58) Energy-consistent
Stuttgart/Cologne
koeln.de/PP/clickpse.en.html.
pseudopotentials
of
the
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
complexes arising from decomposition of Ru-H
2
IMes catalysts,
Group, http://www.tc.uni-
see refs 39, 40.
(42) Foster, J. P.; Weinhold, F. J. Am. Chem. Soc. 1980, 102,
7211–7218.
(59) (a) Feller, D. J. Comput. Chem. 1996, 17, 1571–1586. (b)
Schuchardt, K. L.; Didier, B. T.; Elsethagen, T.; Sun, L.;
Gurumoorthi, V.; Chase, J.; Li, J.; Windus, T. L. J. Chem. Inf.
Model. 2007, 47, 1045–1052.
(43) Wiberg, K. B. Tetrahedron 1968, 24, 1083–1096.
(44) Occhipinti, G.; Jensen, V. R. Organometallics 2011, 30,
3
522–3529.
45) Jung, K.-H.; Kim, H.-K.; Lee, G. H.; Kang, D.-S.; Park,
J.-A.; Kim, K. M.; Chang, Y.; Kim, T.-J. J. Med. Chem. 2011, 54,
(60) Dunning, T. H. J. Chem. Phys. 1989, 90, 1007–1023.
(61) Woon, D. E.; Dunning, T. H. J. Chem. Phys. 1993, 98,
1358–1371.
(
5
385–5394.
46) Bottomley, F.; Mukaida, M. J. Chem. Soc., Dalton Trans.
982, 1933–1937.
47) van Lierop, B. J.; Reckling, A. M.; Lummiss, J. A. M.;
Fogg, D. E. ChemCatChem 2012, 4, 2020–2025.
48) Dubberley, S. R.; Romero, P. E.; Piers, W. E.;
McDonald, R.; Parvez, M. Inorg. Chim. Acta 2006, 359, 2658–
664.
49) Fulmer, G. R.; Miller, A. J. M.; Sherden, N. H.;
(62) Martin, J. M. L.; Bauschlicher, C. W.; Ricca, A. Comput.
Phys. Commun. 2001, 133, 189–201.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
1
(63) Zhao, Y.; Truhlar, D. G. Phys. Chem. Chem. Phys. 2008,
10, 2813–2818.
(
(64) Ribeiro, R. F.; Marenich, A. V.; Cramer, C. J.; Truhlar,
D. G. J. Phys. Chem. B 2011, 115, 14556–14562.
(
(65) Fukui, K. Acc. Chem. Res. 1981, 14, 363–368.
(66) (a) Page, M.; Doubleday, C.; McIver, J. W. J. Chem.
Phys. 1990, 93, 5634–5642. (b) Page, M.; McIver, J. W. J. Chem.
Phys. 1988, 88, 922–935.
2
(
Gottlieb, H. E.; Nudelman, A.; Stoltz, B. M.; Bercaw, J. E.;
Goldberg, K. I. Organometallics 2010, 29, 2176–2179.
(67) (a) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev.
Lett. 1996, 77, 3865–3868. (b) Perdew, J.; Burke, K.; Ernzerhof,
M. Phys. Rev. Lett. 1997, 78, 1396–1396.
(
50) Wang, R.; Groux, L. F.; Zargarian, D. Organometallics
002, 21, 5531–5539.
51) Lucke, A. J.; Young, D. J. J. Org. Chem. 2005, 70, 3579–
3583.
(52) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria,
2
(
(68) Grimme, S.; Ehrlich, S.; Goerigk, L. J. Comput. Chem.
2011, 32, 1456–1465.
(69) Johnson, E. R.; Becke, A. D. J. Chem. Phys. 2005, 123,
024101.
G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.;
Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li,
X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.;
Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.;
Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.;
Nakai, H.; Vreven, T.; Montgomery, J. A.; Peralta, J. E.; Ogliaro,
F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.;
Staroverov, V. N.; Keith, T.; Kobayashi, R.; Normand, J.;
Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.;
Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J.
E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts,
R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.;
Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.;
Dapprich, S.; Daniels, A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J.
V.; Cioslowski, J.; Fox, D. J., Gaussian 09, Revision D.01,
Gaussian, Inc., Wallingford CT. 2013.
(70) Smith, D. G. A.; Burns, L. A.; Patkowski, K.; Sherrill,
C. D. J. Phys. Chem. Lett. 2016, 7, 2197–2203.
(71) Kendall, R. A.; Dunning, T. H.; Harrison, R. J. J. Chem.
Phys. 1992, 96, 6796–6806.
(72) (a) Cossi, M.; Scalmani, G.; Rega, N.; Barone, V. J.
Chem. Phys. 2002, 117, 43–54. (b) Scalmani, G.; Frisch, M. J. J.
Chem. Phys. 2010, 132, 114110. (c) Tomasi, J.; Mennucci, B.;
Cammi, R. Chem. Rev. 2005, 105, 2999–3093. (d) Tomasi, J.;
Persico, M. Chem. Rev. 1994, 94, 2027–2094.
(73) (a) Tomasi, J. Theor. Chem. Acc. 2004, 112, 184–203. (b)
Cramer, C. J.; Truhlar, D. G. Phys. Chem. Chem. Phys. 2009, 11,
10757–10816. (c) Klamt, A.; Mennucci, B.; Tomasi, J.; Barone, V.;
Curutchet, C.; Orozco, M.; Luque, F. J. Acc. Chem. Res. 2009, 42,
489–492.
(74) Glendening, E. D.; Badenhoop, J. K.; Reed, A. E.;
Carpenter, J. E.; Bohmann, J. A.; Morales, C. M.; Landis, C. R.;
Weinhold, F. NBO, 6.0.; Theoretical Chemistry Institute,
University of Wisconsin: Madison, WI, 2013.
(
(
(
53) Becke, A. D. J. Chem. Phys. 1997, 107, 8554–8560.
54) Wu, Q.; Yang, W. J. Chem. Phys. 2002, 116, 515–524.
55) Chai, J.-D.; Head-Gordon, M. Phys. Chem. Chem. Phys.
2
008, 10, 6615–6620.
(56) Minenkov, Y.; Singstad, Å.; Occhipinti, G.; Jensen, V.
R. Dalton Trans. 2012, 41, 5526–5541.
(57) Peterson, K. A.; Figgen, D.; Dolg, M.; Stoll, H. J. Chem.
Phys. 2007, 126, 124101.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 18
TOC graphic
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
18