Quantitative catalyst-substrate association relationships between metathesis molybdenum or ruthenium carbene complexes and their substrates

Article abstract of DOI:10.1021/ja104193s

Herein we present the long-sought quantitative catalyst-substrate association relationships based on experimentally measured quantitative association preferences of diverse metathesis Mo and Ru catalysts (Mo-1, Schrock Mo; Mo-2, Schrock-Hoveyda Mo; Ru-1, Grubbs first generation Ru; Ru-2, Grubbs second generation Ru; Ru-3:, Grubbs-Hoveyda first generation Ru; and Ru-4, Grubbs-Hoveyda second generation Ru) to their substrates (alkenes, alkynes and allenes), determined directly by a general method based on FRET principle. The determined substrate preferences are proved to be dependent on the molecular identity of the catalyst, exhibiting the preference order of alkyne > alkene > allene for Mo-1 and Mo-2, allene > alkene > alkyne for Ru-1 and Ru-3, and alkyne > allene > alkene for Ru-2 and Ru-4. The results enable us to probe metathesis mechanisms by answering issues in metathesis reactions including the controversial reaction initiation in enyne or allenyne metathesis.

Full text of DOI:10.1021/ja104193s

Published on Web 08/10/2010
Quantitative Catalyst-Substrate Association Relationships
between Metathesis Molybdenum or Ruthenium Carbene
Complexes and Their Substrates
Kyung Hwan Kim,† Taedong Ok,‡,§ Kwangyeol Lee,| Hee-Seung Lee,§
Kyu Tae Chang,⊥ Hyotcherl Ihee,*,† and Jeong-Hun Sohn*,‡
Center for Time-ResolVed Diffraction, Department of Chemistry, Graduate School of
Nanoscience & Technology (WCU), KAIST, Daejeon, Korea, Institut Pasteur Korea,
Seongnam, Korea, Molecular-LeVel Interface Research Center, Department of Chemistry,
KAIST, Daejeon, Korea, Department of Chemistry, Korea UniVersity, Seoul, Korea, and The
National Primate Research Center, Korea Research Institute of Bioscience and Biotechnology,
Ochang, Korea
Received May 15, 2010; E-mail: sohnjh@ip-korea.org; hyotcherl.ihee@kaist.ac.kr
Abstract: Herein we present the long-sought quantitative catalyst-substrate association relationships based
on experimentally measured quantitative association preferences of diverse metathesis Mo and Ru catalysts
(Mo-1, Schrock Mo; Mo-2, Schrock-Hoveyda Mo; Ru-1, Grubbs first generation Ru; Ru-2, Grubbs second
generation Ru; Ru-3:, Grubbs-Hoveyda first generation Ru; and Ru-4, Grubbs-Hoveyda second generation
Ru) to their substrates (alkenes, alkynes and allenes), determined directly by a general method based on
FRET principle. The determined substrate preferences are proved to be dependent on the molecular identity
of the catalyst, exhibiting the preference order of alkyne > alkene > allene for Mo-1 and Mo-2, allene >
alkene > alkyne for Ru-1 and Ru-3, and alkyne > allene > alkene for Ru-2 and Ru-4. The results enable
us to probe metathesis mechanisms by answering issues in metathesis reactions including the controversial
reaction initiation in enyne or allenyne metathesis.
species with regioselectivity,9 even the substrate-association
preference of this famous class of catalyst and its dependence
Introduction
The discovery of Mo and Ru carbene complexes by Schrock
and Grubbs in 19901 and 1992,2 respectively, has made rapid
progress in synthetic organic chemistry.3 The Mo and Ru
carbenes are transition metal catalysts which can perform a
variety of metathesis reactions among alkene, alkyne and allene
substrates. The diversity of Mo- and Ru-based metathesis
reactivity is reflected by the presence of numerous metathesis
reactions including olefin metathesis,4 enyne metathesis,5 allene
metathesis,6 diyne metathesis,7 and allenyne metathesis.8
Great advances have been made in terms of synthetic
applications of the metathesis reactions, but there are still
ambiguities in the mechanistic understanding, with the most
salient unresolved fundamental issue being the catalyst-substrate
association relationships. Although the association preference
of a catalyst to a substrate among multiple candidate substrates
dictates the identity of the initial adduct/intermediate that can
determine the reaction pathway and the ultimate key propagating
on the molecular identity of the catalyst have not been
quantitatively determined. In particular, the reaction initiation
in ring closing enyne metathesis has been a long-standing
controversy,10 much of which stems from the identical product
from two equally plausible reaction pathways and the deduction
(4) For recent representative reviews on Ru catalyzed olefin metathesis,
see: (a) Grubbs, R. H. Tetrahedron 2004, 60, 7117. (b) Schrock,
R. R.; Czekelius, C. AdV. Synth. Catal. 2007, 349, 55. (c) Hoveyda,
A. H.; Zhugralin, A. R. Nature 2007, 450, 243. For a special issue
dedicated to olefin metathesis, see: (d) AdV. Synth. Catal. 2007, 349,
issues 1-2.
(5) For recent representative reviews on the enyne metathesis, see: (a)
Mori, M. Materials 2010, 3, 2087. (b) Diver, S. T.; Giessert, A. J.
Chem. ReV. 2004, 104, 1317. (c) Villar, H.; Frings, M.; Bolm, C. Chem.
Soc. ReV. 2007, 36, 55. (d) Hansen, E. C.; Lee, D. Acc. Chem. Res.
2006, 39, 509. (e) Kotha, S.; Meshram, M.; Tiwari, A. Chem. Soc.
ReV. 2009, 38, 2065. (f) Chattopadhyay, S. K.; Karmakar, S.; Biswas,
T.; Majumdar, K. C.; Rahaman, H.; Roy, B. Tetrahedron 2007, 63,
3919. (g) Mulzer, J.; Oehler, E. Top. Organomet. Chem. 2004, 13,
269.
† Graduate School of Nanoscience & Technology (WCU), KAIST.
‡ Institut Pasteur Korea.
(6) (a) Ahmed, M.; Arnauld, T.; Barrett, A. G. M.; Braddock, D. C.; Flack,
K.; Procopiou, P. A. Org. Lett. 2000, 2, 551. (b) Janssen, C. E.; Krause,
N. Eur. J. Org. Chem. 2005, 2005, 2322.
§ Department of Chemistry, KAIST.
| Korea University.
(7) (a) Fox, H. H.; Wolf, M. O.; O’Dell, R.; Lin, B. L.; Schrock, R. R.;
Wrighton, M. S. J. Am. Chem. Soc. 1994, 116, 2827. (b) Schrock,
R. R.; Luo, S.; Lee, J. C. J.; Zanetti, N. C.; Davis, W. M. J. Am.
Chem. Soc. 1996, 118, 3883.
⊥ Korea Research Institute of Bioscience and Biotechnology.
(1) Schrock, R. R.; Murdzek, J. S.; Bazan, G. C.; Robbins, J.; DiMare,
M.; O’Regan, M. J. Am. Chem. Soc. 1990, 112, 3875.
(2) Nguyen, S.-B. T.; Johnson, L. K.; Grubbs, R. H. J. Am. Chem. Soc.
1992, 114, 3974.
(8) Murakami, M.; Kadowaki, S.; Matsuda, T. Org. Lett. 2005, 7, 3953.
(9) For representative reviews, see: (a) Gorin, D. J.; Toste, F. D. Nature
2007, 446, 395. (b) Saito, S.; Yamamoto, Y. Chem. ReV. 2000, 100,
2901. (c) Ojima, I.; Tzamarioudaki, M.; Li, Z.; Donovan, R. J. Chem.
ReV. 1996, 96, 635.
(3) (a) Grubbs, R. H. AdV. Synth. Catal. 2007, 349, 34 (Nobel Lecture
2005). (b) Schrock, R. R. AdV. Synth. Catal. 2007, 349, 41 (Nobel
Lecture 2005).
9
10.1021/ja104193s  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 12027–12033 12027

A R T I C L E S
Kim et al.
Figure 1. Quantitative substrate-association preferences of Mo and Ru catalysts for metathesis reaction. (A) Schematic illustration of probing the substrate-
association preference of metathesis Mo and Ru carbene complexes. Mo carbenes (Mo-1 and Mo-2) or Ru carbenes (Ru-1, Ru-2, Ru-3 and Ru-4) act as
quenchers of the coumarin or dapoxyl fluorophore, respectively, when dye-conjugated substrate is bound to the catalyst. (B) Black dashed/dotted lines are
the UV-vis spectra of Mo-1 and Mo-2 (20 µM), and red line is the fluorescence spectrum of the dye molecule in toluene. (C) Black dashed/dotted lines are
the UV-vis spectra of four Ru catalysts (20 µM) and colored solid lines are the fluorescence spectra of the dye molecule in CH2Cl2 (red) and n-hexane
(blue).
of the very early event without direct observation. Theoretical
interpretation of the results has been attempted previously,11
but definitive experimental evidence is currently missing.
Due to the dearth of knowledge on the quantitative substrate
preference, it could not be established how much the molecular
identity of diverse metathesis Mo and Ru catalysts influences their
substrate preferences and thus whether each Mo or Ru catalyst
directs different reaction pathways in the presence of multiple
substrates. Herein we report the quantitative catalyst-substrate
association relationships of an exhaustive list of metathesis Mo
and Ru catalysts to various substrates, and discuss their
usefulness in answering previously unexplained observations
in metathesis reactions and resolving the assumption on the
controversial reaction initiation in enyne or allenyne metathesis.
substrates that provide direct and critical information on the
substrate-association preferences of the catalysts and initial
adduct/intermediate formation in the metathesis reactions.
Since all of the two Mo and four Ru catalysts have absorbance
bands at visible range with no emission of fluorescence and
thus could act as fluorescence quenchers, we chose two dyes,
coumarin and dapoxyl dyes that have fluorescence emission
bands near the absorbance bands of the Mo and Ru catalysts,
respectively (Figure 1A). We prepared the substrates 1-8, the
coumarin- and the dapoxyl-conjugated terminal alkenes, alkynes,
allenes and alkanes (controls) by attaching the desired functional
groups to the dyes via an amide linkage with a three-carbon
tether, as shown in Figure 1A. The emission bands of substrates
1-3 in toluene were observed at 412 nm, and those of substrates
5-7 were observed at 508 nm in CH2Cl2 and 432 nm in
n-hexane, respectively (Figure 1B and 1C).
Results and Discussion
The time-dependent fluorescence quenching of substrates
caused by each catalyst under various reaction conditions
allowed for the simultaneous determination of k, k-1, Kd and
∆G of each specific association event occurring in a given
substrate/catalyst pair. Some examples of the time-dependent
fluorescence spectra of the substrates are shown in Figure 2 for
the six catalysts, and the time-dependent fluorescence traces
(integrated values of fluorescence spectra) for all possible
combinations of the six catalysts and the six substrates are
plotted in Figure 3. During the fluorescence quenching, no
significant absorbance change of the substrate due to the energy
transfer between the substrates was observed (see Supporting
Information), confirming that it is the catalyst that is responsible
for the fluorescence quenching.
Our plan for determining the substrate preferences of the Mo
and Ru catalysts entailed bringing fluorescence resonance energy
transfer (FRET) principle into effect (Figure 1).12 Using the
FRET-based method we directly detected the interactions
between the catalysts (Mo-1, Schrock Mo; Mo-2, Schrock-
Hoveyda Mo; Ru-1, Grubbs first generation Ru; Ru-2, Grubbs
second generation Ru; Ru-3, Grubbs-Hoveyda first generation
Ru; and Ru-4, Grubbs-Hoveyda second generation Ru) and
dye-conjugated substrates (alkenes, alkynes and allenes). We
determined both kinetic and thermodynamic parameters for the
association event between the Mo or the Ru complexes and the
(10) (a) Kim, S. H.; Bowden, N.; Grubbs, R. H. J. Am. Chem. Soc. 1994,
116, 10801. (b) Hoye, T. R.; Donaldson, S. M.; Vos, T. J. Org. Lett.
1999, 1, 277. (c) Schramm, M. P.; Reddy, D. S.; Kozmin, S. A. Angew.
Chem., Int. Ed. 2001, 40, 4274. (d) Hansen, E. C.; Lee, D. J. Am.
Chem. Soc. 2003, 125, 9582. (e) Lee, H.-Y.; Kim, B. G.; Snapper,
M. L. Org. Lett. 2003, 5, 1855. (f) Lloyd-Jones, G. C.; Margue, R. G.;
de Vries, J. G. Angew. Chem., Int. Ed. 2005, 44, 7442. (g) Diver,
S. T.; Galan, B. R.; Giessert, A. J.; Keister, J. B. J. Am. Chem. Soc.
2005, 127, 5762. (h) Lee, Y.-J.; Schrock, R. R.; Hoveyda, A. H. J. Am.
Chem. Soc. 2009, 131, 10652.
The dependence of the fluorescence quenching rate on the
substrate is clearly seen for each catalyst. For example, Figure
3A exhibits substrate-dependent, disparate fluorescence quench-
ing rates of the three coumarin-conjugated substrates for catalyst
Mo-1 with λmax ) 364 nm, which is fastest for alkyne 2 with
the decreasing rate order of alkyne 2 > alkene 1 > allene 3. For
each catalyst, the fluorescence quenching traces for the three
substrates under various reaction conditions were simultaneously
(11) Lippstreu, J. J.; Straub, B. F. J. Am. Chem. Soc. 2005, 127, 7444.
(12) Sohn, J.-H.; Kim, K. H.; Lee, H.-Y.; No, Z. S.; Ihee, H. J. Am. Chem.
Soc. 2008, 130, 16506.
9
12028 J. AM. CHEM. SOC. VOL. 132, NO. 34, 2010

Catalyst-Substrate Association Relationships
A R T I C L E S
Figure 2. Time-dependent change of the fluorescence spectra of dye-alkene (1 or 5), dye-alkyne (2 or 6) and dye-allene (3 or 7) due to the reaction with
the (A) Mo-1, (B) Mo-2, (C) Ru-1, (D) Ru-2, (E) Ru-3 and (F) Ru-4 catalysts. For each pair of a catalyst and a substrate, multiple experimental conditions
(for example, 20 µM:20 µM and 10 µM:20 µM of substrate vs catalyst and mixed substrates for some cases) were used to increase the information content
of data, and here the time-dependent spectra for only one condition (20 µM:20 µM) are shown.
Figure 3. Time-dependent fluorescence quenching of the dye-conjugated substrates (alkenes 1 and 5, alkynes 2 and 6, and allenes 3 and 7) by six types of
catalysts, (A) Mo-1, (B) Mo-2, (C) Ru-1, (D) Ru-2, (E) Ru-3, and (F) Ru-4. Experimental data points are shown as black circles and squares (20 µM:20
µM and 10 µM:20 µM of substrate vs catalyst, respectively), and the theoretical curves from the global fitting analysis are represented as red, blue and green
solid lines for alkenes 1 and 5, alkynes 2 and 6, and allenes 3 and 7, respectively. The black circles (20 µM:20 µM) correspond to the integrated values of
the fluorescence spectra shown in Figure 2. Fitting was done by using a simple kinetic model (see Supporting Information), and it elucidates the kinetic and
thermodynamic behavior of the association of the catalyst with the substrate. The orderings of substrate-association preference of each catalyst are shown
at the top of the each panel.
fitted against a set of rate equations (see Supporting Informa-
tion), and the results are shown in Table 1. The obtained rate
coefficients and the Gibbs free energy for the association process
between the substrate and the catalyst provide more quantitative
information for the substrate preference of each catalyst than
can be visualized from Figure 3.
Generally the final product can be reached via various reaction
pathways and each reaction pathway may consist of multiple
reaction steps. The rate determining step is the one with the
highest reaction barrier along a reaction pathway and depends
on the identity of the catalyst used in the reaction. According
to previous studies, for Mo-catalyzed olefin metathesis reaction,
it has been shown that the initial substrate association step is
the rate determining step.13 Thus, our experimentally obtained
substrate preferences can be safely used to rationalize the
suggested mechanisms in Mo-catalyzed metathesis reactions.
For catalyst Mo-1, the substrate preference order of alkyne
> alkene > allene, shown in Figure 3 and Table 1, is in line
with the common assumption that ring closing enyne metathesis
catalyzed by group VI (Cr, Mo, W) complexes initiates at alkyne
(path B in Figure 4A),14 and the exclusive endo-product
formations in the reactions using Mo complexes through
metallacyclobutene intermediates.10h The stronger preference of
Mo-1 to alkyne than allene can also resolve the mechanism of
the ring closing allenyne metathesis using Mo-1, along with
the observation of byproduct A and the complete consumption
of alkyne with remaining allene intact in the attempted
intermolecular allenyne metathesis using Mo-1 (Figure 4B),8
indicating the reaction initiation at alkyne.4b
The Schrock-Hoveyda catalyst (Mo-2) with λmax ) 361 nm
possesses the same order of substrate preference as the Schrock
catalyst (Mo-1) both kinetically and thermodynamically, which
is alkyne 2 > alkene 1 > allene 3 (Figure 3B and Table 1).
Considering their similar electronic and molecular structures,
(14) (a) Katz, T. J.; Sivavec, T. M. J. Am. Chem. Soc. 1985, 107, 737. (b)
Korkowski, P. F.; Hoye, T. R.; Rydberg, D. B. J. Am. Chem. Soc.
1988, 110, 2676. (c) Watanuki, S.; Ochifuji, N.; Mori, M. Organo-
metallics 1994, 13, 4129.
(13) Poater, A.; Solans-Monfort, X.; Clot, E.; Coperet, C.; Eisenstein, O.
J. Am. Chem. Soc. 2007, 129, 8207.
9
J. AM. CHEM. SOC. VOL. 132, NO. 34, 2010 12029

A R T I C L E S
Kim et al.
Table 1. Kinetic and Thermodynamic Parameters for Alkenes 1 and 5, Alkynes 2 and 6, and Allenes 3 and 7 with Each Mo and Ru Catalyst
Determined by the Data Analysisa
catalyst
substrate
k (M-1 s-1
)
rel kb
k-1 (s-1
)
Kd (k-1/k)
∆G (kJ/mol)
rel stabilityc
Mo-1
1
2
3
1
2
3
5
6
7
5
6
7
5
6
7
5
6
7
6.54((2.51) × 102
1.81((0.67) × 103
6.07((0.87) × 102
8.75((3.07) × 102
1.05((0.34) × 103
4.67((2.15) × 102
3.71((0.13) × 103
2.60((0.09) × 102
5.25((0.53) × 104
1.19((0.33) × 102
2.98((0.34) × 102
1.63((0.27) × 102
4.34((5.29) × 101
2.55((2.62) × 101
1.19((0.29) × 102
1.10((0.53) × 102
8.63((0.42) × 102
4.98((0.87) × 102
1
3.45((1.45) × 10-3
2.50((0.88) × 10-3
3.56((1.41) × 10-3
1.28((1.95) × 10-5
1.52((2.71) × 10-6
2.57((3.53) × 10-5
2.12((0.23) × 10-2
9.81((8.15) × 10-3
8.71((1.59) × 10-3
1.98((1.39) × 10-2
5.26((4.05) × 10-4
2.79((0.33) × 10-3
6.96((9.31) × 10-2
5.64((8.72) × 10-2
1.25((0.91) × 10-2
1.62((1.78) × 10-2
1.31((4.67) × 10-5
1.45((0.63) × 10-2
5.28((3.00) × 10-6
1.39((0.70) × 10-6
5.87((2.50) × 10-6
1.46((2.29) × 10-8
1.46((2.63) × 10-9
5.51((7.95) × 10-8
5.71((0.65) × 10-6
3.77((3.14) × 10-5
1.66((0.35) × 10-7
1.66((1.26) × 10-4
1.77((1.37) × 10-6
1.71((0.35) × 10-5
1.60((2.90) × 10-3
2.21((4.11) × 10-3
1.05((0.80) × 10-4
1.47((1.77) × 10-4
1.52((5.41) × 10-8
2.91((1.36) × 10-5
-29.6 ( 1.38
-32.9 ( 1.23
-29.3 ( 1.03
-43.9 ( 3.78
-49.6 ( 4.44
-40.7 ( 3.53
-29.4 ( 0.03d
-24.8 ( 2.03d
-38.0 ( 0.05
-21.9 ( 1.90
-33.4 ( 1.96
-27.6 ( 0.51
-16.2 ( 4.56
-15.4 ( 4.68
-23.1 ( 1.93
-22.2 ( 3.02
-45.4 ( 8.98
-26.3 ( 1.18
1
3.80
0.90
1
10.0
0.26
1
0.15
34.4
1
93.8
9.70
1
2.77
0.93
1
1.2
0.53
1
0.07
14.2
1
2.50
1.37
1
0.59
2.60
1
7.85
4.53
Mo-2
Ru-1
Ru-2
Ru-3
Ru-4
d
d
d
d
0.72
15.2
1
9670
5.05
a k and k-1 are directly determined as fitting parameters, and Kd and ∆G are calculated with proper equations, k-1/k and -RT ln(1/Kd), repectively. k
and ∆G represent the kinetic and theromodynamic preferences respectively. b Relative speed of the association reaction. The k of alkene (1 or 5) is
defined as 1. c Relative thermodynamic stability of the associated complex. The Kd of alkene-catalyst complex is defined as 1. d Values referred from
our previous study.12
> alkyne 6,12 both kinetically and thermodynamically (Figure
3C and Table 1). Although allene metathesis reaction has not
drawn much attention from synthetic chemists and, to our best
knowledge, only two bisallene metathesis6 and one allenyne
metathesis8 have been reported, our results may stimulate intense
research on the metathesis reactions with allene substrate as
well as allene transformation reactions catalyzed by other
transition metal catalysts.
Different from the case of Ru-1, the time-dependent fluo-
rescence quenching by Ru-2 with λmax ) 499 nm is fastest and
most favorable for alkyne, followed by allene and then by alkene
(alkyne 6 > allene 7 > alkene 5), caused by the effect of the
imidazolidine carbene ligand of Ru-2 in the place of one
phosphine (PCy3) of Ru-1 (Figure 3D and Table 1). Our results
are in accord with the previous suggestion by Grubbs and co-
workers that Ru-1 and Ru-2 possess different phosphine binding
and dissociation behaviors that result in slower binding rate of
alkene to Ru-2 than Ru-1.15 Our results also, for the first time,
quantitatively show that catalysts Ru-2, Mo-1 and Mo-2
commonly favor alkyne over alkene or allene, and directly
differentiate Ru-2 from Ru-1 by their completely different
substrate preferences.
Figure 4. Some examples of metathesis reaction patterns using Mo carbene
that have been previously unexplained but can be explained in light of the
determined substrate-association preference ordering. (A) The substrate
preference order of alkyne > alkene > allene is in line with the common
assumption that ring closing enyne metathesis catalyzed by group VI
complexes initiates at alkyne (path B).14 (B) Ring-closing allenyne
metathesis catalyzed by Mo-1. Along with the observation of byproduct A
and the complete consumption of alkyne with remaining allene intact in
the attempted intermolecular allenyne metathesis using Mo-1,8 the stronger
preference of Mo-1 to alkyne than to allene can resolve the mechanism,
indicating the reaction initiation at alkyne.4b
The Grubbs-Hoveyda catalysts (Ru-3 and Ru-4) differ from
the Grubbs analogues (Ru-1 and Ru-2) in that the empty
coordination site generated by the dissociation of a PCy3 ligand
is lightly protected by a weakly coordinating oxygen atom of
an ether moiety. Considering the weak coordination nature of
the oxygen atom, Ru-3 and Ru-4 should exhibit similar substrate
preferences to their Ru-1 and Ru-2 analogues, respectively.
Catalysts Ru-3 and Ru-4 exhibit small absorbance bands at 500
and 580 nm in CH2Cl2, respectively, which barely overlap with
the fluorescence of the dye in the same solvent, not suitable for
FRET study. Therefore we changed the reaction medium to
n-hexane so that the fluorescence band of the substrates is shifted
to 432 nm and overlaps with the huge absorbance bands of Ru-3
and Ru-4 at 363 and 377 nm, respectively (Figure 1C).
the same preference order is anticipated although some differ-
ences are observed for the relative fluorescence quenching rates
between the substrates, which might be due to the different
electronic effect between alkoxides and phenoxides of the
catalysts. This is the first time that the exact ordering of the
substrate preference has been quantitatively determined for both
Mo-1 and Mo-2 with all their metathesis substrates, alkene,
alkyne and allene.
Unlike Schrock and Schrock-Hoveyda catalysts, the Grubbs
first generation catalyst (Ru-1) with λmax ) 527 nm possesses
completely reverse substrate preference, allene 7 > alkene 512
(15) Sanford, M. S.; Ulman, M.; Grubbs, R. H. J. Am. Chem. Soc. 2001,
123, 749.
9
12030 J. AM. CHEM. SOC. VOL. 132, NO. 34, 2010

Catalyst-Substrate Association Relationships
A R T I C L E S
Figure 5. Summary of the substrate preferences of Mo and Ru catalysts studied. The substrate preferences are dependent on the molecular identity of the
catalyst, exhibiting the preference order of (A) alkyne > alkene > allene for Mo-1 and Mo-2, (B) allene > alkene > alkyne for Ru-1 and Ru-3, and (C) alkyne
> allene > alkene for Ru-2 and Ru-4. Based on their favorite substrates, these catalysts can be categorized into two groups, namely, Mo-1, Mo-2, Ru-2 and
Ru-4 favoring alkyne, and Ru-1 and Ru-3 favoring allene (A and B vs C).
As expected, catalyst Ru-3 exhibits the same substrate
association preference of allene 7 > alkene 5 > alkyne 6 as for
Ru-1, and catalyst Ru-4 demonstrates the substrate preference
of alkyne 6 > allene 7 > alkene 5, identical to that for Ru-2
(Figure 3E,F and Table 1). In n-hexane, catalyst Ru-1 with
another absorbance at 330 nm also displayed the same substrate
preference order as in CH2Cl2, confirming the general ap-
plicability of our FRET-based method (see Supporting Informa-
tion). While some differences in the fluorescence quenching rate
were observed, overall reactivity patterns remain the same.
Unlike the case of Mo-catalyzed olefin metathesis in which
the rate determining step is the substrate association step, for
the Ru case, low temperature NMR studies provided results
supporting that the rate determining step is ancillary ligand
decoordination for olefin metathesis,16 and some quantum
calculations support that the highest energy point is at a
metallacyclobutene for enyne metathesis.11 Thus, the obtained
substrate-association preferences cannot be directly applied to
answer previously unexplained Ru-catalyzed reaction patterns.
On the other hand, it is rather surprising that some reaction
patterns appear to be linked with the substrate association
preferences although our experiment alone cannot justify the
assumption that the initial substrate association determines the
final outcome. For example, the observation that bisallene cross-
metathesis reactions proceed well by Ru-1, but not by Ru-2,6a
is in line with the relatively strong allene substrate preference
of Ru-1 compared with that of Ru-2. Additionally, much
stronger alkyne preference of Ru-2 than to alkene substrate can
be in accordance with the result that only the alkyne group was
transformed by Ru-2 whereas the alkene group remains intact
in the trimerization reaction of cyclooctenyne, an enyne cyclic
compound with similarly distorted alkene and alkyne.17 It should
be noted, however, that although the substrate preferences appear
to direct the reaction outcome, the current experimental data
concerning only the initial substrate-catalyst association step
alone cannot prove such dependence. To obtain the whole
perspective, each plausible downstream process needs to be
investigated, and in principle, this can be achieved by designing
FRET pairs pertaining to such a step and using the same FRET-
based method.18
catalyst, exhibiting the preference order of alkyne > alkene >
allene for Mo-1 and Mo-2 presumably due to the similar
electronic and molecular structure of the two catalysts, allene
> alkene > alkyne for Ru-1 and Ru-3, and alkyne > allene >
alkene for Ru-2 and Ru-4. Considering the weak coordination
nature of the ether oxygen atom, Ru-3 and Ru-4 should exhibit
similar substrate preferences to their Ru-1 and Ru-2 analogues,
respectively. Based on their favorite substrates, these catalysts
can be categorized into two groups, namely, Mo-1, Mo-2, Ru-2
and Ru-4 favoring alkyne, and Ru-1 and Ru-3 favoring allene
(Figure 5). To fully understand the intricate play of energetics
between substrates and catalysts, further theoretical studies on
molecular orbital energy, as well as the investigation of the steric
effect of the substrate, are necessary.
Conclusion
We quantitatively determined the catalyst-substrate relation-
ships with association preferences of the representative Mo and
Ru catalysts for metathesis reaction to their substrates, alkene,
alkyne and allene using FRET-based method. The results
provide new insight into the metathesis reactions along with
answering the reaction initiation in enyne or allenyne metathesis
and previously unexplained metathesis patterns. Due to our
successful quantification of the Mo/Ru catalyst-substrate as-
sociation interactions, diverse metal complex-substrate systems
can now be quantitatively assessed and the newly found set of
the catalyst-substrate association relationships might lead to
new metal catalysis such as new tandem or multicomponent
reaction methodologies, aided by much clarified reaction
mechanisms.
Experimental Section
1. Synthesis of Compounds.
1.1. General. Common solvents were purified before use.
Tetrahydrofuran (THF) and dichloromethane (CH2Cl2) were purified
by distillation from sodium-benzophenone and calcium hydride
respectively. Methanol and Et3N (AldrichSeal) were used as
received. All reagents were reagent grade and purified where
necessary. “Water” refers to distilled water. Reactions were
monitored by thin layer chromatography (TLC) using Whatman
precoated silica gel plates. Flash column chromatography was
performed over ultra pure silica gel (230-400 mesh) from Merck.
1H NMR and 13C NMR spectra were recorded on Varian 400-MR
spectrometer and Bruker AVANCE 400 spectrometer using residual
solvent peaks as an internal standard (CHCl3, δ 7.24 ppm for proton
and δ 77.0 ppm for carbon; acetone, δ 2.05 ppm for proton and δ
In summary, our results demonstrate that the substrate
preferences are dependent on the molecular identity of the
(16) van der Eide, E. F.; Piers, W. E. Nat. Chem. 2010, 2, 571.
(17) Carness, M.; Buccella, D.; Siegrist, T.; Steigerwald, M. L.; Nuckolls,
C. J. Am. Chem. Soc. 2008, 130, 14078.
(18) Lim, S.-G.; Blum, S. A. Organometallics 2009, 28, 4643.
9
J. AM. CHEM. SOC. VOL. 132, NO. 34, 2010 12031

A R T I C L E S
Kim et al.
29.9 ppm for carbon; CH3OH, δ 3.30 ppm for proton and δ 49.0
ppm for carbon). Multiplicities for 1H NMR are designated as: s )
singlet, d ) doublet, dd ) doublet of doublets, ddd ) doublet of
dd, dt ) doublets of triplets, dq ) doublet of quartets, td ) triplet
of doublets, t ) triplet, q ) quartet, m ) multiplet, bs ) broad
singlet. Infrared spectra (IR) were recorded on a Bruker ALPHA-E
FT-IR spectrometer and are reported in reciprocal centimeters
(cm-1). UV-visible spectra were recorded on Varian Cary 5000
spectrophotometer. High resolution mass spectra (HRMS) were
obtained on Bruker microTOF-Q.
2.30 (td, J ) 7.1 Hz, 2.6 Hz, 1H), 2.12 (td, J ) 6.6 Hz, 2.6 Hz,
1H), 1.96 (t, J ) 2.6 Hz, 0.5H), 1.88 (m, 1H), 1.78 (m, 1H), 1.67
(t, J ) 2.6 Hz, 0.5H). 13C NMR (100 MHz, CD3OD): δ 165.6,
165.5, 163.7, 163.5, 158.5, 158.0, 156.2, 156.0, 143.3, 142.9,
129.51, 129.46, 122.1, 121.9, 113.3, 113.2, 111.99, 111.96, 100.61,
100.59, 83.6, 82.7, 69.3, 68.8, 55.9, 49.6, 46.9, 38.6, 36.7, 32.7,
26.6, 25.8, 15.9, 15.5. IR (film) cm-1: 3435, 3284, 2935, 2848,
1717, 1611, 1565, 1508, 1441, 1404, 1369, 1292, 1243, 1196, 1027,
836, 637, 525. UV/vis: λmax 345 nm. Fluorescence: λmax 412 nm
(toluene). HRMS (m/z): [M + Na]+ calcd for C17H17NNaO4,
322.1055; found, 322.1061.
1.2. General Procedure for the Syntheses of Coumarin-
1.5. Coumarin-Allene 3 (56%). TLC (n-hexane:EtOAc, 40:
60 v/v): Rf ) 0.30. 1H NMR (400 MHz, CD3OD): δ 7.83 (s, 0.5H),
7.76 (s, 0.5H), 7.39(d, J ) 8.6 Hz, 1H), 6.87 (dt, J ) 8.6 Hz, 2.1
Hz, 1H), 6.81 (d, J ) 2.3 Hz, 1H), 5.14 (m, 0.5H), 4.97 (m, 0.5H),
4.68 (m, 1H), 4.52 (m, 1H), 3.86 (s, 3H), 3.53 (t, J ) 7.3 Hz, 1H),
3.28 (t, J ) 7.4 Hz, 1H), 3.05 (s, 1.5H), 2.96 (s, 1.5H), 2.09 (m,
1H), 1.88 (m, 1H), 1.70 (m, 2H). 13C NMR (100 MHz, CD3OD):
δ 208.5, 208.3, 165.4, 165.3, 163.6, 163.6, 158.4, 158.0, 156.1,
156.0, 143.1, 142.6, 129.5, 129.4, 122.3, 122.1, 113.2, 112.0, 111.9,
100.64, 100.61, 89.3, 88.8, 75.6, 75.3, 55.9, 50.3, 47.2, 36.4, 32.8,
27.0, 26.1, 25.2, 24.7. IR (film) cm-1: 3419, 2923, 2852, 1953,
1714, 1606, 1562, 1506, 1440, 1402, 1367, 1290, 1241, 1192, 1147,
1116, 1021, 798, 611, 520. UV/vis: λmax 345 nm. Fluorescence:
λmax 412 nm (toluene). HRMS (m/z): [M + Na]+ calcd for
C18H19NNaO4, 336.1212; found, 336.1216.
Alkene 1, -Alkyne 2, -Allene 3, and -Alkane 4.
To a stirred solution of aldehyde (S1,19 S220 or S321) (110 mg,
1.34 mmol) in MeOH was added MeNH2 (33 wt % in EtOH, 0.33
mL, 2.68 mmol), and the mixture was stirred at room temperature
for 2 h. After the reaction mixture was cooled to 0 °C, NaBH4 (25
mg, 0.67 mmol) was added and the resulting mixture was stirred
at 0 °C for 30 min. After quenching with H2O (3 mL) the mixture
was extracted with CH2Cl2 (5 mL × 3), dried over MgSO4, filtered,
acidified with 2 M HCl in Et2O (2 mL), and concentrated under
reduced pressure. The solid crude product was washed with Et2O
to afford off-white solid. To a stirred solution of the solid (18.5
mg, 0.14 mmol), coumaric acid (15 mg, 0.068 mmol), HOBT (2
mg, 0.014 mmol), and HATU (52 mg, 0.14 mmol) in THF (0.22
mL) was added i-Pr2NEt (0.060 mL, 0.34 mmol), and the mixture
was stirred overnight at room temperature in a dark atmosphere.
Saturated NH4Cl (1 mL) was added, and the resulting mixture was
extracted with CH2Cl2 (1 mL × 3). The combined organic layers
were dried over MgSO4, filtered, and concentrated under reduced
pressure. The crude product was purified by prep. TLC (EtOAc:
n-hexane ) 6:4) to give coumarin-conjugated compound as
yellowish film-like solid.
1.6. Coumarin-Alkane 4. 4 was synthesiszed from com-
mercially available N-methylbutylamine (88%). TLC (n-hexane:
1
EtOAc, 40:60 v/v): Rf ) 0.32. H NMR (400 MHz, CD3OD): δ
7.82 (s, 0.5H), 7.76 (s, 0.5H), 7.39 (d, J ) 8.6 Hz, 1H), 6.84 (dt,
J ) 8.6 Hz, 2.1 Hz, 1H), 6.80 (d, J ) 2.3 Hz, 1H), 3.86 (s, 3H),
3.49 (t, J ) 7.4 Hz, 1H), 3.23 (t, J ) 7.5 Hz, 1H), 3.04 (s, 1.5H),
2.95 (s, 1.5H), 1.61 (m, 1H), 1.56 (m, 1H), 1.40 (m, 1H), 1.20 (m,
1H), 0.94 (t, J ) 7.3 Hz, 1.5H), 0.82 (t, J ) 7.3 Hz, 1.5H). 13C
NMR (100 MHz, CD3OD): δ 165.2, 163.6, 158.0, 156.1, 143.0,
142.5, 129.5, 129.4, 122.4, 122.2, 113.2, 112.0, 111.9, 100.7, 100.6,
55.9, 50.9, 47.5, 36.3, 32.7, 30.2, 29.0, 20.0, 19.7, 13.9, 13.7. IR
(film) cm-1: 3384, 2957, 2923, 2853, 1719, 1608, 1563, 1457, 1403,
1367, 1242, 1218, 1158, 1115, 1023, 801, 519. UV/vis: λmax 345
nm. Fluorescence: λmax 412 nm (toluene). HRMS (m/z): [M + Na]+
calcd for C16H20NNaO4, 312.1212; found, 312.1213.
1.7. Synthesis of Dapoxyl-Allene 7.
1.3. Coumarin-Alkene 1 (51%). TLC (n-hexane:EtOAc, 40:
60 v/v): Rf ) 0.31. 1H NMR (400 MHz, CD3OD): δ 7.83 (s, 0.5H),
7.76 (s, 0.5H), 7.39 (d, J ) 8.7 Hz, 1H), 6.85 (dt, J ) 8.6 Hz, 2.3
Hz, 1H), 6.80 (d, J ) 1.9 Hz, 1H), 5.82 (m, 0.5H), 5.64 (m, 0.5H),
5.05 (d, J ) 17.1 Hz, 0.5H), 4.97 (d, J ) 10.3 Hz, 0.5H), 4.92 (d,
J ) 17.1 Hz, 0.5H), 4.86 (d, J ) 10.2 Hz, 0.5H), 3.86 (s, 3H),
3.50 (t, J ) 7.4 Hz, 1H), 3.24 (t, J ) 7.5 Hz, 1H), 3.04 (s, 1.5H),
2.95 (s, 1.5H), 2.12 (q, J ) 7.0 Hz, 1H), 1.94 (q, J ) 6.9 Hz, 1H),
1.69 (m, 2H). 13C NMR (100 MHz, CD3OD): δ 165.4, 165.3, 163.6,
158.4, 158.0, 156.1, 156.0, 143.2, 142.7, 137.8, 137.1, 129.5, 129.4,
122.3, 122.0, 115.5, 115.1, 113.2, 112.0, 111.9, 100.64, 100.60,
A mixture of S422 (500 mg, 2.20 mmol) and H2NNH2 ·H2O (80%
in H2O, 0.7 mL) in THF (3 mL) was stirred overnight at 40 °C and
for 5 h at room temperature. The reaction mixture was filtered and
rinsed with Et2O (20 mL). The filtrate was washed with 1 N NaOH
(4 mL) and brine (4 mL), dried over MgSO4, filtered, acidified with
2 M HCl in Et2O (3 mL), and concentrated under reduced pressure.
The solid crude product was washed with EtOAc and CH2Cl2 to
give hexa-4,5-dien-1-aminium chloride23 as colorless solid (168
mg, 57%): 1H NMR (400 MHz, CD3OD) δ 5.18 (m, 1H), 4.73 (m,
2H), 2.96 (t, J ) 7.6 Hz, 2H), 2.09 (m, 2H), 1.77 (m, 2H); 13C
NMR (100 MHz, CD3OD) δ 208.7, 88.0, 74.6, 39.0, 26.6, 24.7.
To a mixture of the hexa-4,5-dien-1-aminium chloride (1.2 mg,
9.05 µmol) and activated dapoxyl (2.5 mg, 6.17 µmol) in CH2Cl2
55.9, 50.4, 47.3, 36.3, 32.7, 30.8, 30.4, 27.1, 26.1. IR (film) cm-1
:
3424, 2924, 2853, 1717, 1607, 1563, 1507, 1441, 1403, 1367, 1291,
1241, 1196, 1157, 1137, 1117, 1024, 912, 833, 620. UV/vis: λmax
345 nm. Fluorescence: λmax 412 nm (toluene). HRMS (m/z): [M +
Na]+ calcd for C17H19NNaO4, 324.1212; found, 324.1206.
1.4. Coumarin-Alkyne 2 (62%). TLC (n-hexane:EtOAc, 40:
60 v/v): Rf ) 0.31. 1H NMR (400 MHz, CD3OD): δ 7.84 (s, 0.5H),
7.78 (s, 0.5H), 7.39 (d, J ) 8.6 Hz, 1H), 6.85 (dt, J ) 8.6 Hz, 2.1
Hz, 1H), 6.80 (t, J ) 2.7 Hz, 1H), 3.86 (s, 3H), 3.60 (t, J ) 7.1
Hz, 1H), 3.40 (t, J ) 7.1 Hz, 1H), 3.05 (s, 1.5H), 2.99 (s, 1.5H),
(19) Kelly, B. D.; Allen, J. M.; Tundel, R. E.; Lambert, T. H. Org. Lett.
2009, 11, 1381.
(20) Adams, T. C.; Dupont, A. C.; Carter, J. P.; Kachur, J. F.; Guzewska,
M. E.; Rzeszotarski, W. J.; Farmer, S. G.; Noronha-Blob, L.; Kaiser,
C. J. Med. Chem. 1991, 34, 1585.
(22) Trost, B. M.; Pinkerton, A. B.; Seidel, M. J. Am. Chem. Soc. 2001,
123, 12466.
(21) Tsukamoto, H.; Matsumoto, T.; Kondo, Y. J. Am. Chem. Soc. 2008,
130, 388.
(23) Jonasson, C.; Horva´th, A.; Ba¨ckvall, J.-E. J. Am. Chem. Soc. 2000,
122, 9600.
9
12032 J. AM. CHEM. SOC. VOL. 132, NO. 34, 2010

Catalyst-Substrate Association Relationships
A R T I C L E S
(0.2 mL) was added Et3N (3.8 µL, 27.3 µmol), and the resulting
mixture was stirred for 20 min at 0 °C. The mixture was diluted
with EtOAc (3 mL), washed with saturated NH4Cl (0.5 mL) and
brine (0.5 mL), dried over MgSO4, filtered, and concentrated under
reduced pressure. The crude product was purified by preparative
TLC to give dapoxyl-allene 7 (2.3 mg, 95%) as a pale green solid.
TLC (n-hexane:EtOAc, 30:70 v/v): Rf ) 0.18. 1H NMR (400 MHz,
acetone-d6): δ 8.15 (d, J ) 8.4 Hz, 2H), 8.02 (d, J ) 8.4 Hz, 2H),
7.87 (bs, 1H), 7.68 (d, J ) 8.8 Hz, 2H), 7.45 (s, 1H), 6.84 (d, J )
8.8 Hz, 2H), 5.20 (m, 1H), 4.69 (m, 2H), 3.47 (q, J ) 6.0 Hz, 2H),
3.02 (s, 6H), 2.12 (m, 2H), 1.76 (m, 2H);. 13C NMR (100 MHz,
acetone-d6): δ 209.5, 206.1, 166.6, 159.7, 153.8, 151.8, 137.1, 130.9,
128.7, 126.6, 126.4, 122.0, 116.6, 113.2, 90.2, 75.3, 40.4, 40.1,
26.5. IR (film) cm-1: 3329, 2924, 2854, 2360, 2341, 1956, 1728,
1638, 1611, 1509, 1364, 1288, 1125, 853, 815, 714. UV/vis: λmax
366 nm. Fluorescence: λmax 566 nm (MeOH), 548 nm (EtOH), 545
nm (CH3CN), 530 nm (i-PrOH), 521 nm (acetone), 508 nm
(CH2Cl2), 488 nm (EtOAc), 454 nm (m-xylene), 432 nm (n-hexane).
HRMS (m/z): [M + H]+ calcd for C24H26N3O2, 388.2025; found,
388.2024.
2.2. Measurement of Fluorescence Quenching Signal. The
time-dependent fluorescence quenching signal was measured by a
home-built spectrofluorometer consisting of a Nd:YAG nanosecond
laser, a grating and a CCD detector (for more details, see Supporting
Information). During the fluorescence measurement, all the samples
were treated with anhydrous solvents under Ar. A 3 mL solution
of a substrate in toluene (for Mo-1 and Mo-2), CH2Cl2 (for Ru-1
and Ru-2) or n-hexane (for Ru-3 and Ru-4) in a quartz cuvette
with Ar balloon was put in a temperature-controlled holder. The
sample was excited at 355 nm, and the fluorescence spectrum was
recorded from 320 to 600 nm for toluene solutions, 350 to 650 nm
for CH2Cl2 solutions, and 360 to 560 nm for n-hexane solutions.
Before a catalyst was added, a spectrum was acquired to provide
a reference spectrum at time zero. 1 or 2 equiv of the catalyst (0.3
mM CH2Cl2 solution) was added to the substrate solution using a
25 µL syringe, and the fluorescence spectrum was obtained as a
function of time.
Acknowledgment. We thank Hee-Yoon Lee (KAIST), Tae-
Lim Choi (SNU) and JeanSun Lim for helpful discussions on
metathesis reaction or experimental assistance. This work was
partially supported by Creative Research Initiatives (Center for
Time-Resolved Diffraction) of MEST/NRF and the NRF grant
(No.2010-0001953) funded by MEST. We acknowledge the support
from the WCU program.
2. Measurement of Time-Dependent Fluorescence Quench-
ing Signal.
2.1. Home-Built Spectrometer. Time-dependent fluorescence
quenching signal was measured by a home-built spectrofluorometer
that consists of a Nd:YAG nanosecond laser, a grating and a CCD
detector. The third harmonic (355 nm) of an Nd:YAG laser
(Brilliant B) was used as the excitation beam. Laser pulses with
10 µJ were focused to about 5 mm diameter. Sample solution in
10 by 10 mm quartz cuvette was put in a temperature-controlled
holder. Fluorescence light was gathered by two lenses located at
90° from the laser path and spatially resolved by wavelength with
the grating. Resolved fluorescence spectrum was recorded from 350
to 650 nm by a gated CCD detector. The measurement speed was
10 Hz, and typically 150 or 300 (corresponding to 15 or 30 s)
spectra were averaged to get a time-dependent spectrum at a certain
time point.
Supporting Information Available: All time-dependent quench-
ing traces and their conditions, substrate absorbance change
during fluorescence quenching, time-dependent fluorescence
quenching of the substrates by Ru-1 in n-hexane, global fitting
analysis of FRET data, and 1H and 13C NMR spectra of
substrates 1-4 and 7. This material is available free of charge
via the Internet at http://pubs.acs.org.
JA104193S
9
J. AM. CHEM. SOC. VOL. 132, NO. 34, 2010 12033