Ruthenium Olefin Metathesis Catalysts Containing Fluoride

Article abstract of DOI:10.1021/acscatal.5b00219

The reaction of the ruthenium complex cis-Caz-1 with silver fluoride affords the first example of an active olefin metathesis precatalyst containing fluoride ligands. The cis geometry of the precursor complex is key to the successful fluoride exchange rea

Full text of DOI:10.1021/acscatal.5b00219

Research Article
pubs.acs.org/acscatalysis
Ruthenium Olefin Metathesis Catalysts Containing Fluoride
Stefano Guidone,† Olivier Songis,† Laura Falivene,‡ Fady Nahra,† Alexandra M. Z. Slawin,†
Heiko Jacobsen,§ Luigi Cavallo,‡ and Catherine S. J. Cazin*,†
†EaStCHEM School of Chemistry, University of St Andrews, St. Andrews KY16 9ST, United Kingdom
‡Physical Sciences and Engineering, Catalysis Center, King Abdullah University of Science and Technology (KAUST), Thuwal
23955-6900, Saudi Arabia
§KemKom, 1215 Ursulines Avenue, New Orleans, Louisiana 70116, United States
S
* Supporting Information
ABSTRACT: The reaction of the ruthenium complex cis-Caz-
1 with silver fluoride affords the first example of an active
olefin metathesis precatalyst containing fluoride ligands. The
cis geometry of the precursor complex is key to the successful
fluoride exchange reaction. Computational studies highlight
the stability of the new Ru−F species, due to push−pull
interactions between fluoride and L-type ligands (L: N-
heterocyclic carbene, phosphite). Insights into the isomer-
ization process from trans-Caz-1 to cis-Caz-1 are given.
Fluoride exchange reactions were performed involving cis-
and trans-Caz-1 complexes. Catalytic tests showcase the
excellent activity of the Ru−F containing complexes.
KEYWORDS: olefin metathesis, homogeneous catalysis, ruthenium, fluoride, halide exchange
INTRODUCTION
■
The introduction of fluorine atoms in molecules can drastically
change their physical and chemical properties.1 Organofluorine
compounds are nowadays employed in the agrochemical and
pharmaceutical industry, with an ever increasing demand for this
class of compounds.2 Fluorine has been introduced into metal
complexes as a fluoride ligand.3 Late-transition-metal fluoride
complexes represent a challenge in synthetic organometallic
chemistry because of the hard−soft mismatch.2,4 As a result, only
a limited number of examples of transition metal fluoride
complexes (namely, Pd,5a Pt,5b Ir,5c Os,5d−f and Ru5e−k) have
been disclosed to date. Most synthetic protocols leading to M−F
bond formation involve an oxidative addition of XeF2 in
anhydrous HF or reaction with Et3N·3HF.4d,f,5d,6 Oxidative
addition of organofluoro compounds via C−F bond activation
can be performed, but it is somewhat limited to group 9−10
metals.7 An alternative strategy to access M−F compounds
consists of halide-exchange reactions that can be carried out
under relatively mild reaction conditions using reagents such as
CsF,4c,d AgF,4f NMe4F,3h,4g and [(Me2N)3S]+[Me3SiF2]−
(TASF).4h Such synthetic methods have been applied to
Figure 1. Ruthenium fluoride complexes.
ruthenium, leading to fluoride complexes, as shown in Figure
1. Caulton and co-workers have reported the synthesis of
[Ru(F)(H)(CO)(PtBu2Me)2] through halide exchange reac-
adopted with the fluoride ligand trans to the π-acceptor CO
(Figure 1). A push−pull interaction between the fluoride (here
tions using CsF.4c More recently, Whittlesey and co-workers
have shown that complexes of type [Ru(F)(H)(CO)Ln] (n = 2,3;
L = PR3, NHC) could be isolated using Et3N·3HF as the
fluorinating agent.8 In all complexes reported by Caulton and
Whittlesey, a square pyramidal or an octahedral geometry is
Received: February 3, 2015
Revised: May 13, 2015
© XXXX American Chemical Society
3932
DOI: 10.1021/acscatal.5b00219
ACS Catal. 2015, 5, 3932−3939

ACS Catalysis
Research Article
a
acting as a π-donor) and the trans carbonyl ligand is invoked to
stabilize the Ru−F bond.
Table 1. Chloride Exchange with AgF
Difluoride Ru clusters such as [RuF2(CO)3]45i−k developed by
Wilson and co-workers and derivatives obtained by substitution
of carbonyls with phosphines5f and N-heterocyclic carbenes
(NHC)5h were reported by Hope and co-workers. The reactivity
of the 18e−-difluoride complexes [RuF2(CO)2L2] (L = PR3,
NHC) toward Lewis acids (LA) (e.g., PF5, BF3, and B(C6F5)3)
has led to the corresponding 16e−-cationic species, in which the
counterion has a stabilizing effect on the complex.5g,h All reports
on Ru-carbonyl complexes contrast with the cationic mono-
fluoride and the neutral difluoride species developed by Mezzetti
and co-workers (Figure 1).3g,h This last cationic complex was
obtained through halide exchange of [RuCl(dppp)2]PF6 with
TlF (dppp = 1,3-bis(diphenylphosphino)propane), and further
reacted with NMe4F to yield the neutral difluoride species. In
these complexes, the fluoride ligand is described as a poor π-
donor ligand, resulting in a simple electrostatic interaction
between a cationic ruthenium species and fluoride. To the best of
our knowledge, no other type of Ru complex bearing a F ligand
has been reported to date. Complexes that would be of great
interest would be Ru−F-containing congeners of alkene
metathesis active systems. Indeed, kinetic studies by Grubbs
and co-workers showed that the catalytic activity of complexes of
the type [RuX2(SIMes)(CHPh)] (X = Cl, Br, I, SIMes =
N,N′-bis[2,4,6-(trimethyl)phenyl]imidazolidin-2-ylidene) de-
creases in the order Cl > Br > I, suggesting a possibly higher
efficiency for the fluoride analogues.9 Computational studies on
the formation of ruthena(IV)cyclobutanes from first- and
second-generation Grubbs precatalysts and norborn-2-ene have
illustrated a similar trend of reactivity for dihalide complexes (F >
Cl > Br > I), predicting the best activity of difluoride species
among olefin metathesis precatalysts.10 Despite this insight into
possible reactivity, whereas bromide/iodide/carboxylate ver-
sions of Grubbs and Hoveyda−Grubbs type complexes have
been extensively studied,9,11 no fluoride version has, to the best of
our knowledge, been reported to date. A few failed attempts
using HF or AgF suggest that these methodologies or Grubbs/
Hoveyda−Grubbs type complexes are not suitable for the
synthesis of Ru−F species for olefin metathesis.12 Recently, we
have developed a new class of Ru-olefin metathesis catalysts that
display an unusual cis geometry and lead to outstanding stability
and activity in olefin metathesis reactions.13 It appeared therefore
of great interest to evaluate the ability of such complexes to
undergo chloride exchange with fluoride. Herein, we report our
findings on chloride exchange reactions using a number of olefin
metathesis precatalysts.
AgF
[Ru]Cl2
b
[Ru]ClF
[Ru]F2
b
b
entry
precatalyst
equiv
species
species
species
1
2
3
4
5
6
7
8
9
Ind-II
1
100
100
100
90
n.r.
n.r.
n.r.
10
n.r.
n.r.
n.r.
Ind-III
1
Hov-II
1
trans-Caz-1
trans-Caz-1
cis-Caz-1
cis-Caz-1
cis-Caz-1
cis-Caz-1
1
c
c
1
47
53
1
24
76
>99
21
1.2
2
79
2.4
>99
a
Reaction conditions: complex (0.01 mmol), AgF (1−2.4 equiv),
b
1
CD2Cl2 (0.6 mL), rt, 1 h. Conversion (%) determined by H NMR.
c
24 h.
1−3). In contrast, under the same reaction conditions, the
phosphite-containing complex trans-Caz-1 leads to 10%
conversion of the starting material into the monofluorinated
species (Table 1, entry 4). With longer reaction time (24 h), 53%
of trans-Caz-1 was converted into a Ru−F species (Table 1, entry
5). Monitoring this reaction for a few hours showed a reaction
rate similar to the one involved in the trans−cis isomerization of
the complex.14 We therefore reasoned that isomerization might
be required before chloride exchange. If such were the case, the
cis complex should show a superior reactivity. Cis-Caz-1 was thus
reacted with AgF. A very rapid reaction occurred with a 76%
conversion leading to the same fluorinated product obtained
from trans-Caz-1 (Table 1, entry 6). Using 2 equiv of AgF, full
conversion of the starting material was observed with both
mono- and difluoride products formed. Using only 20% excess of
AgF leads to the selective formation of the monofluorinated
product (Table 1, entries 7 and 8). Increasing the amount of AgF
to 2.4 equiv leads to the quantitative formation of the difluoride
species (Table 1, entry 9).
Synthesis of [Ru(Cl)(F)(Ind)(SIMes){P(OiPr)3}], Caz-1F.
Caz-1F was quantitatively obtained by reaction of 1 equiv of cis-
Caz-1 with 1.2 equiv of AgF. The 1H NMR spectrum contains a
characteristic doublet at 8.93 ppm assigned to the indenylidene
(Ind) H7 proton, which is shifted downfield compared with the
signal observed for cis-Caz-1. In the 31P−{1H} NMR spectrum, a
doublet is observed at 131.4 ppm (2JPF = 286 Hz) corresponding
to the phosphite ligand. This signal is also shifted downfield
compared with the 31P signal of the starting material.13a In the
19F−{1H} NMR spectrum, a doublet at −217.2 ppm (2JPF = 286
Hz) is found. These data are consistent with a trans disposition of
the fluorine and phosphite ligands. The 13C−{1H} NMR
spectrum contains a doublet at 289.6 ppm (2JCP = 25.5 Hz)
and a doublet at 209.1 ppm (2JCP = 16.0 Hz) corresponding to
the carbene carbon atom of the indenylidene and of the N-
heterocyclic carbene, respectively. The structure was unambig-
uously confirmed by single-crystal X-ray diffraction (Figure 3).15
Caz-1F presents a slightly distorted square pyramidal geometry
with the indenylidene unit sitting at the apical position.
Hydrogen bonding (dashed line) between H7 of the inden-
ylidene ring and the fluorine atom is observed, which might
explain the smaller CInd−Ru−F angle (102.82(13)°) observed
compared with the dichloride analogue (104.3(2)°).13a Similar
intramolecular F···H interactions have also been observed by
Hope and co-workers.5f
RESULTS AND DISCUSSION
■
Chloride Exchange with AgF. The reactivity of various
commercially available precatalysts with AgF as a fluoride source
was initially examined (Figure 2).
The reaction of Ind-II, Ind-III, and Hov-II with AgF showed
no conversion of the starting material after 1 h (Table 1, entries
Figure 2. Olefin metathesis precatalysts tested in Cl exchange.
3933
DOI: 10.1021/acscatal.5b00219
ACS Catal. 2015, 5, 3932−3939

ACS Catalysis
Research Article
Table 2. Selected Bond Distances (Å), Angles (deg)(esd)
Ru(1)−XtransP
Ru(1)−XtransNHC
Ru(1)−P(1)
Ru(1)−C(24)
Ru(1)−C(1)
P(1)−Ru(1)−C(24)
P(1)−Ru(1)−C(1)
C(24)−Ru(1)−C(1)
2.4036(18)
2.3974(19)
2.249(2)
2.067(7)
1.881(8)
100.6(19)
90.5(2)
2.029(3)
2.017(3)
2.035(4)
2.2263(16)
2.053(6)
1.846(6)
97.82(14)
89.95(15)
96.1(3)
2.3837(12)
2.2426(11)
2.059(4)
1.864(5)
97.66(10)
90.19(12)
97.09(17)
98.7(3)
Chloride Exchange and Isomerization Studies. Brad-
dock and co-workers have reported a study using Hoveyda−
Grubbs precatalysts of the type [RuX2(SIMes)(CH−2-
(iPrO)−C6H4)] where X is Cl, Br, CF3CO2, and C2F5CO2.11d
When reacting two complexes containing different pairs of
anionic ligands (1:1 molar ratio), a statistical mixture of 1:2:1 was
obtained, in which a novel complex with mixed anions is in
equilibrium with the starting materials (pseudodegenerate ligand
exchange).11d This encouraged us to investigate the ability of our
dichloro and difluoro species to be involved in such an exchange
(Table 3).
Figure 3. Molecular structure of Caz-1F. All solvent molecules and
hydrogen atoms are omitted for clarity. Thermal ellipsoids are shown at
the 50% probability level.
Synthesis of [RuF2(Ind)(SIMes){P(OiPr)3}], Caz-1F2. The
difluorinated complex Caz-1F2 was quantitatively obtained by
1
reaction of cis-Caz-1 with 2.4 equiv of AgF. The H NMR
spectrum contains a characteristic doublet at 9.02 ppm
corresponding to the indenylidene H7 proton, which is shifted
downfield compared with Caz-1F (8.93 ppm) and cis-Caz-1
(8.87 ppm).13a In the 31P−{1H} NMR spectrum, a doublet at
131.5 ppm (2JPF of 286 Hz) corresponding to the phosphite
ligand is observed. This signal is also shifted downfield compared
with the 31P signals of Caz-1F and cis-Caz-1.13a The 19F-{1H}
NMR spectrum contains a doublet at −217.7 ppm (2JPF = 286
Hz) corresponding to the fluoride trans to the phosphite, and a
broad singlet at −237.2 ppm corresponding to the second
fluoride, which is cis to the phosphite ligand. The 13C−{1H}
NMR was recorded at 223 K. The spectrum contains two broad
signals at 287.0 and 211.1 ppm corresponding to the carbene
carbon atom of the indenylidene and of the N-heterocyclic
carbene, respectively. The structure of the complex was
confirmed by X-ray diffraction on s single crystal (Figure 4).16
Table 3. Fluoride Exchange Reactions
a
entry
complex
time (h)
conversion to Caz-1F (%)
1
2
3
4
Caz-1F2 + cis-Caz-1
0.5
0.5
1.5
24
>99
30
Caz-1F2 + trans-Caz-1
40
>99
a
1
Determined by H NMR.
When the dichloride species cis-Caz-1 was mixed with Caz-
1F2, rapid and quantitative reaction was observed, with the
monofluorinated compound being the only product formed
(Table 3, entry 1), which is in marked contrast with the Braddock
findings. Furthermore, complete fluorination of trans-Caz-1 can
be achieved, requiring a longer reaction time (Table 3, entry 4).
This is of great interest considering that the formation of this
compound from trans-Caz-1 was found problematic when using
the AgF-promoted Cl exchange route (Table 1, entries 4 and 5).
In addition, comparison of the rate of isomerization of trans-Caz-
1 to cis-Caz-1 with the rate of fluorination of trans-Caz-1 by Caz-
1F2 showed that the latter process is faster, which is again in
contrast with the results obtained with AgF.17 These results
prompted us to investigate the fluorination of Ind-II, Ind-III, and
Hov-II using Caz-1F2. Unfortunately, no selective formation of
fluorinated compounds was observed in these instances.18
Computational Studies. Insights into the stability of the Ru
complexes isolated in this work and their corresponding trans
species were obtained using a computational DFT approach
(Table 4).19 As can be seen in Table 4, the energy values are
clearly affected by the relative stability of the free chloride and
free fluoride anions in solution.
Focusing on the cis complexes, the replacement of one
chloride ligand of cis-Caz-1 by one fluoride ligand, leading to
Caz-1F + Cl−, is favored by 18.9 kcal/mol (Table 4). In
agreement with the experiments, the most stable isomer of Caz-
1F presents the fluoride ligand trans to the P(OiPr)3 ligand. The
isomer with the fluoride ligand trans to the NHC ligand is 11.6
kcal/mol higher in energy. Substitution of the second chloride
atom in the trans isomer, that is, the one with the P(OiPr)3 ligand
trans to the NHC ligand, is less favored than the substitution of
Figure 4. Molecular structure of Caz-1F2. All solvent molecules and
hydrogen atoms are omitted for clarity. Thermal ellipsoids are shown at
the 50% probability level.
A similar interaction (dashed line) between H7 of indenylidene
ring and the fluorine atom trans to the phosphite is observed,
presumably responsible for the relatively small CInd−Ru−F angle
(103.39(17)°) compared with the dichloride analogue
(104.3(2)°).13a Both complexes display a distorted square
pyramidal geometry with the phosphite ligand cis to the NHC
(P−Ru−CNHC angles being 97.66(10)° and 97.82(14)° for Caz-
1F and Caz-1F2, respectively) (Table 2).
In both complexes, the Ru−F bond is significantly shorter than
the Ru−Cl. For all complexes, comparable Ru−C (NHC and
indenylidene) bond lengths are found, whereas Ru−P bond
distances show some disparity. The latter decreases in the
following order: cis-Caz-1 > Caz-1F > Caz-1F2, which reflects the
electronegativity of F vs Cl.
3934
DOI: 10.1021/acscatal.5b00219
ACS Catal. 2015, 5, 3932−3939

ACS Catalysis
Research Article
the Ru−Cl bond is stronger when it is trans to another Cl atom
rather than to a P atom (comparing entries 2 and 5 in Table 5).
To gauge if the increased strength of the Ru−F bond relative
to the Ru−Cl bond is electrostatic in nature or is due to increased
back-donation from the F ligand to the Ru center through a
push−pull effect induced by the π-acid phosphite, we performed
an energy decomposition analysis (EDA) of the Ru−X bond in
cis-Caz-1 and cis-Caz-1F2 (see Table 6). We briefly remind the
Table 4. Energetics (kcal/mol) of the Chloride Exchange
Reactions
isomer
cis
Caz-1 + 2F−
Caz-1F + F− + Cl−
Caz-1F2 + 2Cl−
0.0
+6.8
+6.8
−18.9
−7.3
−32.6
−17.9
+14.7
trans
trans−cis
+11.6
Table 6. Decomposition Analysis of the Bond-Snapping
Energy of Selected Ru−X Bonds in the Gas Phase
the second chloride in the cis isomer (see Table 4); however, it is
worth noting that in Caz-1, the cis isomer is only 6.8 kcal/mol
more stable than the trans isomer, whereas in Caz-1F and Caz-
1F2, this preference increases to 11.6 and 14.7 kcal/mol,
respectively. Finally, combination of the energy values of the
three cis species in Table 4 indicates that the formation of 2 mol
of Caz-1F by mixing 1 mol of cis-Caz-1 and 1 mol of Caz-1F2 is
favored by 5.2 kcal/mol. Overall, these results are consistent with
the experimental data showing that when cis-Caz-1 is reacted
with 1 equiv of AgF, only Caz-1F is formed, and when cis-Caz-1 is
reacted with cis-Caz-1F2, only Caz-1F is still obtained.
a
EEl
EPauli
EOrb
BSE
cis-Caz-1
−154.7
−168.3
−168.0
117.5
122.7
119.2
−77.9
−92.6
−94.0
−115.0
−138.1
−142.8
cis-Caz-1F2
cis-Caz-1F
a
The total bond-snapping energy is decomposed as −BSE = EEl
EPauli + EOrb. All values in kcal/mol.
+
reader that the total BSE can be decomposed into three main
terms: −BSE = EEl + EPauli + EOrb. EEl accounts for stabilizing
electrostatic interaction between the two fragments, EPauli
accounts for repulsion between doubly filled molecular orbitals
on the two fragments, and EOrb accounts for stabilizing
interaction between filled orbitals on one fragment with empty
orbitals on the other fragment. Because the EDA is performed on
the intrinsic strength of the Ru−X bond, which does not depend
on the environment that can only stabilize the two fragments, the
EDA is performed in the gas-phase. Focusing on cis-Caz-1 and
cis-Caz-1F2, the data reported in Table 6 indicate that the
increased strength of the Ru−F bond in cis-Caz-1F2 compared
with the Ru−Cl bond in cis-Caz-1 is almost 50%−50% shared
between the orbital and the electrostatic terms because both are
roughly 14−15 kcal/mol stronger in cis-Caz-1F2 compared with
cis-Caz-1. Neither of the two terms is clearly dominant. The
repulsive Pauli term is instead rather similar. This suggests that
the increased strength of the Ru−F bond is equally a
consequence of an increased ionic character of the bond, due
to the higher electronegativity of the F atom, and of an increased
orbital interaction, which can be probably correlated to back-
donation from the F ligand to the Ru center through a push−pull
effect. Comparison between cis-Caz-1F2 and cis-Caz-1F indicates
that the halide trans to the NHC ligand has a small impact on the
Ru−F bond trans to the phosphite.
Catalytic Studies in Ring-Closing Metathesis (RCM).
The catalytic activities of both Caz-1F2 and Caz-1F were first
examined in the RCM of the challenging substrate N,N-bis(2-
methylallyl)tosylamide 1, and compared with the catalytic
activity of the dichloro analogue cis-Caz-1. Because thermal
activation is required to efficiently promote such a reaction, a first
comparative study at different temperatures was carried out
(Figure 5).13,21 As can be seen in Figure 5, although the
monofluoride performs very well, the difluoride analogue leads to
poor conversion to 2 (max. 33% at 110 °C). Comparison of cis-
Caz-1 with its monofluoride analogue Caz-1F shows that,
although the catalytic activity of the dichloride derivative reaches
its maximum at 90 °C (79%), the activity of the monofluoride
derivative carries on increasing with temperature (84% at 120
°C). This result is even more striking if we compare the catalytic
profile of the ring-closing reaction of the three complexes at 110
°C (Figure 6). Under such conditions, cis-Caz-1 has a rapid rate
of reaction, followed by a rapid catalyst deactivation (87%
conversion within 15 min), whereas Caz-1F has an induction
A better understanding of the relative strength of the different
Ru−halide bonds in these systems was achieved with bond-
snapping-energy calculations (BSE) which is the energy required
for the dissociation of the Ru−X bond (see Table 5). In these
Table 5. Ru−X Bond-Snapping Energy in CH2Cl2 (kcal/mol)
precatalyst
atom
heterolytic BSE homolytic BSE
1
2
cis-Caz-1
Cl trans to NHC
Cl trans to P
Cl trans to NHC
F trans to P
31.5
21.7
30.9
40.6
36.0
47.5
47.2
37.4
43.5
40.3
79.0
84.3
cis-Caz-1
3
cis-Caz-1F
85.1
4
cis-Caz-1F
110.4
92.8
5
trans-Caz-1
trans-Caz-1F2
trans-Caz-1F
trans-Caz-1F
cis-Caz-1F2
cis-Caz-1F2
equivalent Cl
equivalent F
6
116.1
114.1
97.2
7
F trans to Cl
Cl trans to F
F trans to NHC
F trans to P
8
9
110.2
113.7
10
calculations, the geometry of the [Ru] fragment is kept rigid, and
we calculated both homolytic and heterolytic BSEs, which means
fragmenting the complex into radical and neutral [Ru]· and X·
fragments or into cationic [Ru]+ and anionic X− fragments,
respectively. The BSEs in Table 5 indicate that all the Ru−F BSEs
are stronger than the BSE for the corresponding Ru−Cl bond,
independently from homo or heterolytic fragmentation of the
complexes. This allows us to conclude that the Ru−F bond is
stronger than the corresponding Ru−Cl bond. More specifically,
the data reported in Table 5 clearly indicate that the Ru−F bonds
in cis-Caz-1F2 are roughly 12−18 kcal/mol stronger than the
corresponding Ru−Cl bonds in cis-Caz-1. Moreover, for the
same complexes, the Ru-halide bond trans to the P(OiPr)3 ligand
is from 3 to 10 kcal/mol weaker than the Ru-halide bond trans to
the NHC ligand (comparing entries 9 and 10 as well as entries 1
and 2 in Table 5), suggesting that it is the Ru-halide bond trans to
the P atom that undergoes dissociation. Focusing on cis-Caz-1F,
the Ru−Cl bond is weaker than the Ru−F bond (comparing
entries 3 and 4 in Table 5), despite the chloride ligand being trans
to the NHC ligand. Finally, the strength of the Ru−Cl bond trans
to the NHC ligand is nearly the same in cis-Caz-1 and cis-Caz-1F
(comparing entries 1 and 3 in Table 5). Focusing on trans-Caz-1,
in agreement with our previous work,20 calculations indicate that
3935
DOI: 10.1021/acscatal.5b00219
ACS Catal. 2015, 5, 3932−3939

ACS Catalysis
Research Article
kcal/mol above the corresponding cis isomer (see Table 4). To
complete this pathway, phosphite dissociation from the trans
isomers costs 6.1, 6.0, and 3.9 kcal/mol (trans-Caz-1, trans-Caz-
1F, and trans-Caz-1F2, respectively). As for the stepwise
activation mechanism, we were not able to locate a stable
geometry for the 14e− intermediate with a halide trans to the
NHC ligand. In all cases, this intermediate collapsed to the most
stable geometry with the two halides trans to each other. This
suggests that phosphite dissociation from the cis isomer triggers,
and is concerted with, the rearrangement of the Ru moiety to the
most stable 14e− intermediate with the vacancy trans to the NHC
ligand. We thus located the transition state for this process, and
we found barriers of 28.8, 35.8, and 37.1 for Caz-1, Caz-1F and
Caz-1F2, respectively. Comparison between the two mechanisms
thus indicates that the most likely mechanism for activation of
the cis complexes consists in the dissociation of the phosphite
with concerted isomerization of the Ru moiety to generate the
14e− intermediate with a vacant coordination position trans to
the Ru atom. Furthermore, the trend in the energy of the
transition states for phosphite dissociation from the cis
complexes, easier for Caz-1 and more difficult for Caz-1F2, is
consistent with the trend observed for the metathesis activity of
Figures 5 and 6. As a remark, we note that we are somewhat
overestimating the barrier for phosphite dissociation for Caz-1F,
which should be closer to the barrier calculated for Caz-1.
Finally, the beneficial effect of using a fluorinated aromatic
solvent was examined for all three complexes (Figure 7). As
Figure 5. Temperature dependence in the RCM investigated.
Figure 6. Reaction profiles of cis-Caz-1, Caz-1F, and Caz-1F2 in the
RCM of 1 (lines are visual aids and not curve fits).
period of ∼3 min and reaches 93% conversion within 40 min. On
the other hand, the difluoro species, Caz-1F2, is poorly active
under such reaction conditions (6% conversion after 2 h, and
21% after 24h).22 It should be noted that cis-Caz-1 outperformed
other commercially available precatalysts (such as Ind-II, Hov-
II, and Ind-III) in the RCM of substrate 1 under similar
conditions,13a,c thus suggesting that Caz-1F could be a serious
contender for this transformation.
To shed light on the formation of the catalytically active
species, further DFT calculations were undertaken. For this
purpose, we investigated the possible mechanism leading to
formation of the classic 14e− intermediate, with the Ru atom
presenting a vacant coordination position trans to the NHC
ligand for substrate coordination. Starting from the cis isomers,
two mechanisms can be envisaged. The first corresponds to
concerted cis-to-trans isomerization of the complex with the
phosphite still coordinated to the ruthenium, followed by
phosphite dissociation. The second corresponds to dissociation
of the phosphite from the cis isomer, followed by isomerization
of the resulting 14e− intermediate, presenting one halide trans to
the NHC ligand, to the most stable 14e− intermediate with a
vacant coordination site trans to the NHC ligand.
Figure 7. RCM properties of Ru−F precatalysts and cis-Caz-1 in toluene
and perfluorotoluene under inert atmosphere.
previously suggested by DFT calculations,23 an aromatic solvent
molecule can stabilize the 14e− species by direct coordination to
the Ru center. The same calculations indicated that fluorinated
aromatic solvent molecules interact remarkably more strongly
than their nonfluorinated counterpart, and this result was
correlated to the higher catalytic activity of Ru catalysts in olefin
metathesis. This is consistent with the results depicted in Figure
7 because in all cases, catalytic activity is improved when the
reaction is carried out in perfluorotoluene.
Scope of the reaction. The catalytic efficiency of the
monofluoride complex was next investigated, in a conventional
solvent (toluene), at 110 °C (Figure 8). Caz-1F proved highly
active at 0.1 mol % for hindered and unhindered malonate and
tosylate derivatives. With 0.5 mol % loading, Caz-1F enables
According to calculations, the barrier for the concerted cis−
trans isomerization of the phosphite coordinated species is 35.0,
42.1, and 42.3 kcal/mol for Caz-1, Caz-1F, and Caz-1F2,
respectively, leading to the trans isomers at 6.8, 11.6, and 14.7
3936
DOI: 10.1021/acscatal.5b00219
ACS Catal. 2015, 5, 3932−3939

ACS Catalysis
Research Article
wtih the hypothetical trans-Ru−F species. Although the search
for a viable synthetic route leading to this species still remains
elusive, the present experimental/computational study clearly
highlights that this goal is, indeed, a worthy one. Efforts in this
direction and in directions leading to longer living and more
active metathesis catalysts are ongoing in our laboratory.
EXPERIMENTAL SECTION
■
[Ru(Cl)F(Ind)(SIMes){P(OiPr)3}], Caz-1F. Under an inert
atmosphere of argon, AgF (52.2 mg, 0.41 mmol) was added to a
solution of cis-Caz-1 (303 mg, 0.34 mmol) in dichloromethane
(10 mL).24 The reaction mixture was stirred for 24 h at room
temperature in the absence of light. The solution was filtered, and
the solvent was removed in vacuo, leading to the product as a
pale-brown solid in a quantitative yield (295 mg, 99%). 1H NMR
3
(CD2Cl2, 400 MHz): δ (ppm) = 0.57 (d, JHH = 5.9 Hz, 3H,
CH−CH3), 0.75 (d, 3JHH = 6.0 Hz, 3H, CH−CH3), 0.90 (d, 3JHH
= 5.8 Hz, 3H, CH−CH3), 1.10 (d, 3JHH = 6.0 Hz, 3H, CH−CH3),
1.38 (d, 3JHH = 6.1 Hz, 3H, CH−CH3), 1.44 (d, 3JHH = 6.2 Hz,
3H, CH−CH3), 1.60 (s, 3H, mesityl CH3), 1.83 (s, 3H, mesityl
CH3), 2.33 (s, 3H, mesityl CH3), 2.51 (s, 3H, mesityl CH3), 2.62
(s, 3H, mesityl CH3), 2.69 (s, 3H, mesityl CH3), 3.26 (m, 1H,
CH−CH3), 3.47−3.58 (m, 1H, carbene H4′), 3.80−4.03 (m, 3H,
carbene H4′H5′), 4.26 (m, 1H, CH−CH3), 4.76 (m, 1H, CH−
CH3), 6.03 (s, 1H, mesityl CH), 6.32 (s, 1H, indenylidene H2),
6.51 (s, 1H, mesityl CH), 7.02 (s, 1H, mesityl CH), 7.04 (s, 1H,
Figure 8. Scope of the reaction employing Caz-1F. Reaction conditions:
Caz-1F (0.1 mol %), substrate (0.25 mmol), toluene (0.5 mL), 3 h, 110
°C, under Ar; average of 2 runs; conversions were determined by GC;
isolated yield in parentheses.
a
b
Reaction in neat substrate; 0.5 mol % Ru, isolated as mixture, NMR
c
1
yield; E/Z ratio determined by H NMR.
3
mesityl CH), 7.15 (d, JHH = 7.2 Hz, 1H, indenylidene H4),
7.29−7.37 (m, 2H, indenylidene H5 and H6), 7.37−7.42 (m, 2H,
indenylidene H10), 7.43−7.49 (m, 1H, indenylidene H11), 7.67
(d, 3JHH = 7.2 Hz, 2H, indenylidene H9), 8.93 (d, 3JHH = 7.1 Hz,
1H, indenylidene H7). 13C−{1H} NMR (CDCl3 100.6 MHz) δ
(ppm) = 17.9 (s, mesityl CH3), 18.9 (s, mesityl CH3), 19.1 (s,
mesityl CH3), 20.5 (s, mesityl CH3), 20.7 (s, mesityl CH3), 21.2
(s, mesityl CH3), 24.1 (s, CH−CH3), 24.2 (s, CH−CH3), 24.35
(s, CH−CH3), 24.4 (s, CH−CH3), 24.5 (s, CH−CH3), 24.6 (s,
CH−CH3), 51.6 (s, carbene C4′H), 51.8 (s, carbene C5′H), 68.8
(d, 2JCP = 11.4 Hz, CH−CH3), 69.9 (d, 2JCP = 9.3 Hz, CH−CH3),
72.6 (bs, CH−CH3), 117.3 (s, indenylidene C4H), 127.5 (s,
indenylidene C9H), 128.6 (s, indenylidene C11H), 129.1 (s,
indenylidene C10H), 129.4 (s, indenylidene C6H), 129.6 (s,
indenylidene C5H), 129.7 (s, mesityl CH), 130.1 (s, mesityl
CH), 130.4 (s, mesityl CH), 130.5 (s, indenylidene C7H), 134.5
(s, CIV), 135.3 (s, CIV), 136.3 (s, CIV), 136.8 (s, CIV), 137.0 (s,
CIV), 138.1 (bs, 2 CIV), 138.7 (s, CIV), 139.0 (s, CIV), 139.8 (d,
3JCP = 14.6 Hz, indenylidene C2), 140.2 (s, CIV), 141.6 (m,
enyne metathesis of 6 as well as the efficient and selective cross-
metathesis of 8 with 9.
CONCLUSION
■
The synthesis of the first Ru−F precatalysts for olefin metathesis
reactions has been presented. cis-Caz-1 reacts with AgF under
mild conditions, rather than HF or equivalent reactants, affording
isolable Ru−F species. Both Caz-1F and Caz-1F2 display the same
cis geometry as their chloride precursor. In analogy to what has
been observed by Caulton and co-workers with carbonyl
complexes, a push−pull interaction involving the fluoride ligand
(π-donor) and L-type ligands (π-acceptor) is invoked to stabilize
the complexes (F−Ru(II)−P(OR)3 bonding). Fluoride ex-
change reactions involving Caz-1F2 with cis-Caz-1 and trans-
Caz-1 were performed. The selectivity toward only one product
(Caz-1F) rather than an equilibrium mixture supports the push−
pull effect in the F−Ru(II)−P(OR)3 bonding as a driving force
for the process. Computational studies confirmed the stability of
these complexes when compared with cis-Caz-1, with Caz-1F2
being the most stable. The stability of the cis precatalysts
influences their catalytic activity. A larger energy barrier was
observed for Caz-1F experimentally and in the computational
study compared with cis-Caz-1. Nevertheless, both precatalysts
display similar reactivity in olefin metathesis. Caz-1F2 is less
active than the former precatalysts because of its higher stability
in the cis form and the very small amount of active species
generated. The effect of fluorinated solvents in catalysis was
investigated. An overall enhancement of the catalytic activity was
observed even for Caz-1F2, supporting the hypothesis of a
stabilization of the NHC-Ru species by the fluorinated solvent
rather than the fluorination of active species during the reaction.
The catalytic results reported here are not as superior as
predicted by the literature reports having computationally
addressed trans-Ru−F species. In the present work, the cis
geometry of the species has an influence on the activity of the
complexes, resulting in different catalytic properties compared
2
indenylidene C7a), 142.9 (s, CIV), 209.1 (d, JCP = 16.0 Hz,
carbene C2′), 289.6 (d, 2JCP = 25.5 Hz, indenylidene C1). 31P−
{1H} NMR (CD2Cl2, 162 MHz) δ (ppm) = 131.4 (d, 2JPF = 286
Hz, P(OiPr)3). 19F−{1H} NMR (CD2Cl2, 282.2 MHz) δ (ppm)
2
= −217.2 (d, JPF = 286 Hz, F). Elem. anal. calcd. for
C45H57ClFN2O3PRu: C, 62.81; H, 6.68; N, 3.26. Found: C,
62.72; H, 6.59; N, 3.16.
[RuF2(Ind)(SIMes){P(OiPr)3}], Caz-1F2. Under inert atmos-
phere, AgF (104 mg, 0.82 mmol) was added to a solution of cis-
Caz-1 (298 mg, 0.34 mmol) in dichloromethane (10 mL).25 The
reaction mixture was stirred for 24 h at room temperature in the
absence of light. The solution was filtered, and the solvent was
removed in vacuo, leading to the product as a pale-brown solid in
1
a quantitative yield (280 mg, 98%). H NMR (CD2Cl2, 400
MHz): δ (ppm) = 0.5−1.5 (m, 18H, CH−CH3), 1.66 (s, 3H,
mesityl CH3), 1.83 (s, 3H, mesityl CH3), 2.37 (s, 3H, mesityl
CH3), 2.40 (s, 3H, mesityl CH3), 2.65 (s, 3H, mesityl CH3), 2.68
(s, 3H, mesityl CH3), 3.46−3.58 (m, 1H, carbene H4′), 3.80−
3937
DOI: 10.1021/acscatal.5b00219
ACS Catal. 2015, 5, 3932−3939

ACS Catalysis
Research Article
4.00 (m, 3H, carbene H4′H5′), 6.01 (s, 1H, mesityl CH), 6.38 (s,
Royal Society (University Research Fellowship to C.S.J.C.) for
financial support, and Umicore for the gift of cis-Caz-1.
1H, indenylidene H2), 6.43 (s, mesityl CH), 7.08 (s, 1H, mesityl
3
CH), 7.09 (s, 1H, mesityl CH), 7.16 (d, JHH = 7.1 Hz, 1H,
3
3
indenylidene H4), 7.30 (dd, JHH = 7.1 Hz, JHH = 7.4 Hz, 2H,
REFERENCES
(1) O’Hagan, D. Chem. Soc. Rev. 2008, 37, 308−319.
(2) Torrens, H. Coord. Chem. Rev. 2005, 249, 1957−1985.
■
3
3
indenylidene H5), 7.35 (dd, JHH = 7.1 Hz, JHH = 7.4 Hz, 2H,
3
3
indenylidene H6), 7.40 (dd, JHH = 7.5 Hz, JHH = 7.5 Hz, 2H,
3
3
indenylidene H10), 7.47 (dd, JHH = 7.2 Hz, JHH = 7.4 Hz, 1H
(3) (a) Pagenkopf, B. L.; Carreira, E. M. Tetrahedron Lett. 1998, 39,
9593−9596. (b) Gauthier, D. R.; Carreira, E. M. Angew. Chem., Int. Ed.
Engl. 1996, 35, 2363−2365. (c) Pagenkopf, B. L.; Carreira, E. M. Chem. -
Eur. J. 1999, 5, 3437−3442. (d) Verdaguer, X.; Lange, U. E. W.; Reding,
M. T.; Buchwald, S. L. J. Am. Chem. Soc. 1996, 118, 6784−6785.
(e) Dorta, R.; Egli, P.; Zurcher, F.; Togni, A. J. Am. Chem. Soc. 1997, 119,
10857−10858. (f) Kruger, J.; Carreira, E. M. J. Am. Chem. Soc. 1998, 120,
837−838. (g) Mezzetti, A.; Becker, C. Helv. Chim. Acta 2002, 85, 2686−
indenylidene H11), 7.67 (d, 3JHH = 7.5 Hz, 2H, indenylidene H9),
9.02 (d, JHH = 7.0 Hz, 1H, indenylidene H7). The isopropyl
3
1
protons were observed clearly at lower temperature: H NMR
(CD2Cl2, 400 MHz, 223 K): δ (ppm) = 3.16 (m, 1H, CH−CH3),
3.50 (m, 1H, CH−CH3), 4.64 (m, 1H, CH−CH3). 13C−{1H}
NMR (CDCl3 100.6 MHz,) δ (ppm) = 17.5 (s, mesityl CH3),
18.1 (s, mesityl CH3), 18.6 (s, mesityl CH3), 20.4 (s, mesityl
CH3), 20.8 (s, mesityl CH3), 21.2 (s, mesityl CH3), 23.6−24.7
(bs, 6 CH−CH3), 51.4 (s, carbene C4′H), 51.8 (s, carbene
C5′H), 117.0 (s, indenylidene C4H), 127.3 (s, indenylidene
C9H), 128.3 (s, indenylidene C11H), 129.0 (s, indenylidene
C10H), 129.1 (s, indenylidene C6H), 129.3 (s, indenylidene
C5H), 129.5 (s, mesityl CH), 130.1 (s, mesityl CH), 130.2 (s,
mesityl CH), 130.3 (s, indenylidene C7H), 135.1 (s, CIV), 135.7
(s, CIV), 136.5 (s, CIV), 136.7 (s, CIV), 137.1 (s, CIV), 137.9 (bs, 2
2703. (h) Barthazy, P.; Stoop, R. M.; Worle, M.; Togni, A.; Mezzetti, A.
̈
Organometallics 2000, 19, 2844−2852.
(4) (a) Pearson, R. G. J. Am. Chem. Soc. 1963, 85, 3533−3539.
(b) Caulton, K. G. New. J. Chem. 1994, 18, 25−41. (c) Poulton, J. T.;
Sigalas, M. P.; Folting, K.; Streib, W. E.; Eisenstein, O.; Caulton, K. G.
Inorg. Chem. 1994, 33, 1476−1485. (d) Doherty, N. M.; Hoffman, N. W.
Chem. Rev. 1991, 91, 553−573. (e) Kiplinger, J. L.; Richmond, T. G.;
Osterberg, C. E. Chem. Rev. 1994, 94, 373−431. (f) Murphy, E. F.;
Murugavel, R.; Roesky, H. W. Chem. Rev. 1997, 97, 3425−3468.
(g) Coalter, J. N.; Huffman, J. C.; Streib, W. E.; Caulton, K. G. Inorg.
Chem. 2000, 39, 3757−3764. (h) Burch, R. R.; Harlow, R. L.; Ittel, S. D.
Organometallics 1987, 6, 982−987.
3
CIV), 139.0 (s, CIV), 139.1 (s, CIV), 140.2 (d, JCP = 14.3 Hz,
indenylidene C2), 140.4 (s, CIV), 141.3 (m, indenylidene C7a),
142.1 (s, CIV). Characteristic peaks were observed at lower
temperature (more complicated spectrum due to limited
rotation of the ligands): 13C−{1H} NMR (CDCl3 100.6 MHz,
223 K) δ (ppm) = 68.4 (d, 2JCP = 10.2 Hz, CH−CH3), 69.4 (d,
(5) (a) Braun, T.; Steffen, A.; Schorlemer, V.; Neumann, B.; Stammler,
H. G. Dalton Trans. 2005, 3331−3336. (b) Yahav, A.; Goldberg, I.;
Vigalok, A. Inorg. Chem. 2005, 44, 1547−1553. (c) Cooper, A. C.;
Folting, K.; Huffman, J. C.; Caulton, K. G. Organometallics 1997, 16,
505−507. (d) Brewer, S. A.; Holloway, J. H.; Hope, E. G. J. Chem. Soc.,
Dalton Trans. 1994, 1067−1071. (e) Brewer, S. A.; Coleman, K. S.;
Fawcett, J.; Holloway, J. H.; Hope, E. G.; Russell, D. R.; Watson, P. G. J.
Chem. Soc., Dalton Trans. 1995, 1073−1077. (f) Coleman, K. S.;
Fawcett, J.; Holloway, J. H.; Hope, E. G.; Russell, D. R. J. Chem. Soc.,
Dalton Trans. 1997, 3557−3562. (g) Coleman, K. S.; Fawcett, J.;
Harding, D. A. J.; Hope, E. G.; Singh, K.; Solan, G. A. Eur. J. Inorg. Chem.
2010, 2010, 4130−4138. (h) Fawcett, J.; Harding, D. A. J.; Hope, E. G.;
Singh, K.; Solan, G. A. Dalton Trans. 2009, 6861−6870. (i) Coleman, K.
S.; Holloway, J. H.; Hope, E. G. J. Chem. Soc., Dalton Trans. 1997, 1713−
1718. (j) Marshall, C. J.; Peacock, R. D.; Russell, D. R.; Wilson, I. L. J.
Chem. Soc. D 1970, 1643−1644. (k) Hewitt, A. J.; Holloway, J. H.;
Peacock, R. D.; Raynor, J. B.; Wilson, I. L. J. Chem. Soc., Dalton Trans.
1976, 579−583.
2
2JCP = 7.4 Hz, CH−CH3), 72.0 (bs, JCP = 3.6 Hz, CH−CH3),
211.1 (m, carbene C2′), 287.0 (m, indenylidene C1). 31P−{1H}
NMR (CD2Cl2, 162 MHz) δ (ppm) = 131.5 (d, 2JPF = 286 Hz,
P(OiPr)3). 19F−{1H} NMR (CD2Cl2, 376.3 MHz) δ (ppm) =
−237.2 (br.s, FtransNHC), −217.7 (d, 2JPF = 286 Hz, FtransPhosphite).
Elem. anal. calcd for C45H57F2N2O3PRu: C, 64.04; H, 6.81; N,
3.32. Found: C, 64.03; H, 6.89; N, 3.39.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acscatal.5b00219
(6) Fraser, S. L.; Antipin, M. Y.; Khroustalyov, V. N.; Grushin, V. V. J.
Am. Chem. Soc. 1997, 119, 4769−4770.
(7) Amii, H.; Uneyama, K. Chem. Rev. 2009, 109, 2119−2183.
(8) (a) Reade, S. P.; Nama, D.; Mahon, M. F.; Pregosin, P. S.;
Whittlesey, M. K. Organometallics 2007, 26, 3484−3491. (b) Reade, S.
P.; Mahon, M. F.; Whittlesey, M. K. J. Am. Chem. Soc. 2009, 131, 1847−
1861.
Procedures for catalysis, NMR spectra of complexes and
catalysis products, procedures and NMR spectra for
fluoride exchange reactions and isomerization study
(PDF)
Crystallographic data and structure refinement for
complexes Caz-1F and Caz-1F2, computational details
(CIF); crystallographic data for complexes Caz-1F and
Caz-1F2 in CIF format can also be obtained free of charge
from the Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif (CCDC/911524-
911525)
(9) Sanford, M. S.; Love, J. A.; Grubbs, R. H. J. Am. Chem. Soc. 2001,
123, 6543−6554.
(10) Naumov, S.; Buchmeiser, M. R. J. Phys. Org. Chem. 2008, 21, 963−
970.
(11) (a) Seiders, T. J.; Ward, D. W.; Grubbs, R. H. Org. Lett. 2001, 3,
3225−3228. (b) Krause, J. O.; Nuyken, O.; Wurst, K.; Buchmeiser, M.
R. Chem. - Eur. J. 2004, 10, 777−784. (c) Halbach, T. S.; Mix, S.; Fischer,
D.; Maechling, S.; Krause, J. O.; Sievers, C.; Blechert, S.; Nuyken, O.;
Buchmeiser, M. R. J. Org. Chem. 2005, 70, 4687−4694. (d) Tanaka, K.;
AUTHOR INFORMATION
Corresponding Author
*Fax: (+)44 1334 463808. E-mail: cc111@st-andrews.ac.uk.
■
Bohm, V. P. W.; Chadwick, D.; Roeper, M.; Braddock, D. C.
̈
Organometallics 2006, 25, 5696−5698.
(12) (a) Gawin, R.; Grela, K. Eur. J. Inorg. Chem. 2012, 2012, 1477−
1484. (b) Caskey, S. R.; Stewart, M. H.; Ahn, Y. J.; Johnson, M. J. A.;
Rowsell, J. L. C.; Kampf, J. W. Organometallics 2007, 26, 1912−1923.
(13) (a) Bantreil, X.; Schmid, T. E.; Randall, R. A. M.; Slawin, A. M. Z.;
Cazin, C. S. J. Chem. Commun. 2010, 46, 7115−7117. (b) Schmid, T. E.;
Bantreil, X.; Citadelle, C. A.; Slawin, A. M. Z.; Cazin, C. S. J. Chem.
Commun. 2011, 47, 7060−7062. (c) Songis, O.; Slawin, A. M. Z.; Cazin,
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors gratefully acknowledge the EC through the seventh
framework program (Grant CP-FP 211468-2 EUMET), the
3938
DOI: 10.1021/acscatal.5b00219
ACS Catal. 2015, 5, 3932−3939

ACS Catalysis
Research Article
C. S. J. Chem. Commun. 2012, 48, 1266−1268. (d) Bantreil, X.; Poater,
A.; Urbina-Blanco, C. A.; Bidal, Y. D.; Falivene, L.; Randall, R. A. M.;
Cavallo, L.; Slawin, A. M. Z.; Cazin, C. S. J. Organometallics 2012, 31,
7415−7426. (e) Urbina-Blanco, C. A.; Bantreil, X.; Wappel, J.; Schmid,
T. E.; Slawin, A. M. Z.; Slugovc, C.; Cazin, C. S. J. Organometallics 2013,
32, 6240−6247. (f) Leitgeb, A.; Wappel, J.; Urbina-Blanco, C. A.;
Strasser, S.; Wappl, C.; Cazin, C. S. J.; Slugovc, C. Monatsh. Chem. 2014,
145, 1513−1517.
(14) See Supporting Information, section 3 page S10.
(15) CCDC-911524 contains the supporting crystallographic data for
Caz-1F. These data can be obtained free of charge from the Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/
cif.
(16) CCDC-911525 contains the supporting crystallographic data for
Caz-1F2. These data can be obtained free of charge from the Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/
cif.
(17) See Supporting Information section 5, page S12.
(18) See Supporting Information section 5.2, page S13.
(19) See computational details in the Supporting Information.
(20) Falivene, L.; Poater, A.; Cazin, C. S. J.; Slugovc, C.; Cavallo, L.
Dalton Trans. 2013, 42, 7312−7317.
(21) Slugovc, C.; Perner, B.; Stelzer, F.; Mereiter, K. Organometallics
2004, 23, 3622−3626.
(22) See Supporting Information section 6.3, page S16.
(23) Samojłowicz, C.; Bieniek, M.; Pazio, A.; Makal, A.; Wozniak, K.;
Poater, A.; Cavallo, L.; Wojcik, J.; Zdanowski, K.; Grela, K. Chem. - Eur. J.
2011, 17, 12981−12993.
(24) The granulometry of AgF has an impact on the efficiency of the
chloride exchange reaction in such heterogeneous conditions; see ref 6.
3939
DOI: 10.1021/acscatal.5b00219
ACS Catal. 2015, 5, 3932−3939