Enantio- and regioselective epoxidation of olefinic double bonds in quinolones, pyridones, and amides catalyzed by a ruthenium porphyrin catalyst with a hydrogen bonding site

Article abstract of DOI:10.1021/ja305890c

An array of differently substituted 3-alkenylquinolones was synthesized, and the enantio- and regioselectivity of their Ru-catalyzed epoxidation were studied. A precursor ruthenium(II) complex with a chiral tricyclic γ-lactam skeleton (octahydro-1H-4,7-methanoisoindol-1-one) was available by Sonogashira cross-coupling with a monobromo-substituted ruthenium(II) porphyrin. Enantioselective epoxidation reactions (60-83% yield, 85-98% ee) were achieved with this catalyst, and it was shown that the enantioselectivity depends critically on the presence of a two-point hydrogen bond interaction between the γ-lactam site of the catalyst and the δ-lactam (quinolone) site of the substrate. DFT calculations support the hypothesis that the reaction occurs via a hydrogen-bound transition state, in which the 3-alkenylquinolone adopts an s-trans conformation. The calculations further revealed that this transition state is preferred over a competing s-cis transition state because it exerts less strain in the rigid backbone and because the hydrogen bond interaction is more stable. The catalyst loading required for complete conversion was low (<0.2 mol %), and turnover numbers exceeding 4000 were recorded. It was shown that there is little, if any, inhibition of the catalytic process by other quinolones, which could potentially compete with the binding site. A mechanistic model for the catalytic reaction is presented. In accordance with this model 3-alkenylpyridones reacted with similar enantioselectivities as the respective quinolones. The epoxidation products were unstable, however, and the enantiomeric purity (77-87% ee) of the products could be established only after derivatization. Primary alkenoic acid amides also underwent the epoxidation but gave the respective products in lower enantioselectivities (70% and 45% ee), presumably because the enantioface differentiation is hampered by the increased flexibility of the substrates, which exhibit two or three rotatable single bonds between the binding site and the reactive olefinic double bond.

Full text of DOI:10.1021/ja305890c

Article
pubs.acs.org/JACS
Enantio- and Regioselective Epoxidation of Olefinic Double Bonds in
Quinolones, Pyridones, and Amides Catalyzed by a Ruthenium
Porphyrin Catalyst with a Hydrogen Bonding Site
Philipp Fackler, Stefan M. Huber, and Thorsten Bach*
Lehrstuhl fur Organische Chemie I and Catalysis Research Center (CRC), Technische Universitat Munchen, D-85747 Garching,
̈
̈
̈
Germany
S
* Supporting Information
ABSTRACT: An array of differently substituted 3-alkenylquinolones
was synthesized, and the enantio- and regioselectivity of their Ru-
catalyzed epoxidation were studied. A precursor ruthenium(II)
complex with a chiral tricyclic γ-lactam skeleton (octahydro-1H-4,7-
methanoisoindol-1-one) was available by Sonogashira cross-coupling
with a monobromo-substituted ruthenium(II) porphyrin. Enantiose-
lective epoxidation reactions (60−83% yield, 85−98% ee) were
achieved with this catalyst, and it was shown that the enantioselectivity
depends critically on the presence of a two-point hydrogen bond
interaction between the γ-lactam site of the catalyst and the δ-lactam
(quinolone) site of the substrate. DFT calculations support the hypothesis that the reaction occurs via a hydrogen-bound
transition state, in which the 3-alkenylquinolone adopts an s-trans conformation. The calculations further revealed that this
transition state is preferred over a competing s-cis transition state because it exerts less strain in the rigid backbone and because
the hydrogen bond interaction is more stable. The catalyst loading required for complete conversion was low (<0.2 mol %), and
turnover numbers exceeding 4000 were recorded. It was shown that there is little, if any, inhibition of the catalytic process by
other quinolones, which could potentially compete with the binding site. A mechanistic model for the catalytic reaction is
presented. In accordance with this model 3-alkenylpyridones reacted with similar enantioselectivities as the respective
quinolones. The epoxidation products were unstable, however, and the enantiomeric purity (77−87% ee) of the products could
be established only after derivatization. Primary alkenoic acid amides also underwent the epoxidation but gave the respective
products in lower enantioselectivities (70% and 45% ee), presumably because the enantioface differentiation is hampered by the
increased flexibility of the substrates, which exhibit two or three rotatable single bonds between the binding site and the reactive
olefinic double bond.
and which have been successfully used to control the
regioselectivity of a transition-metal-catalyzed reaction. Im-
portant examples of the use of bifunctional ligands in
regioselective reactions10 include the Ru-catalyzed hydration
of alkynes,11 the Mn-catalyzed C−H oxidation of ibuprofen and
other carboxylic acids,12 and the Rh-catalyzed hydroformylation
of unsaturated carboxylic acids.13 Enantioselectivity in supra-
molecular transition-metal-catalyzed reactions14 has so far been
mediated by hydrogen bonds only in larger entities, in which a
single hydrogen bond is part of the ligand−substrate complex.
An early example was presented by Gilbertson et al., who
studied the Rh-catalyzed hydrogenation of certain olefins in the
presence of peptides, which contained covalently bound
phosphanes.15 Reek et al. have shown the importance of
hydrogen bonding in the enantioselective Rh-catalyzed hydro-
genation with a leucine-derived phosphoramidite ligand
(LEUPhos)16 and with chiral amino acid derivatives as
cofactors of achiral bisphosphane ligands.17
INTRODUCTION
■
Selectivity in transition-metal-catalyzed reactions is achieved by
stabilization of a given transition state relative to other
transition states, which lead to undesired reaction pathways.
The complex nature of substrate and ligand coordination to a
transition metal center1 offers different approaches to selective
transition metal catalysis. Among those approaches, the term
supramolecular catalysis2 describes the idea of using a
supramolecular assembly3 to influence or control the reactivity
of one or more catalytically active centers. More specifically,
supramolecular interactions in transition metal catalysis can be
divided into ligand−ligand and ligand−substrate interactions.4,5
Ligand−substrate interactions enable an indirect communica-
tion of the transition metal center with the substrate via the
ligand. The ligand is crucial for selectivity because its
supramolecular interaction with the substrate is responsible
for the stabilization of a defined transition state. Since the
ligand must bind to both the transition metal and the substrate
it must be at least bifunctional.6 Inspired by natural
precedence,7,8 hydrogen bonds are useful noncovalent inter-
actions,9 which enable supramolecular ligand−substrate binding
Received: June 18, 2012
Published: July 23, 2012
© 2012 American Chemical Society
12869
dx.doi.org/10.1021/ja305890c | J. Am. Chem. Soc. 2012, 134, 12869−12878

Journal of the American Chemical Society
Article
The use of hydrogen bonds as control elements in
enantioselective reactions has been triggered in our group by
the search for noncovalent interactions, which are compatible
with photochemical reaction conditions and which can provide
a sufficient steric bias for enantioface differentiation.18 This
work led to chiral sensitizer 1 (Figure 1), which was shown to
Figure 3. Structure of ruthenium porphyrin complexes 4−6, obtained
by Sonogashira cross-coupling reactions.
compound 4 were synthesized. Compound 4 exhibited
excellent catalytic properties in the enantioselective epoxida-
tion34 of 3-vinylquinolone (95% ee with a 1 mol % catalyst
loading), and by comparison to catalyst 5, it could be shown
that the enantioselectivity is due to the presence of two
hydrogen bonds between the quinolone and the lactam. The
absolute configuration of the resulting epoxide was proven, and
a model was suggested to explain the observed selectivity.
Moreover, it was shown with 3,7-divinylquinolone that the
epoxidation proceeded with higher regioselectivity than the
reaction, which was catalyzed by catalyst 6. We have now
collected further data, which substantiates that the enantiose-
lective epoxidiation is generally applicable to 3-alkenylquino-
lones (10 examples, 86−98% ee). The required starting
materials were synthesized, and their epoxidation reactions
were studied. Previous calculations, which had been conducted
on a semiempirical level, have been refined. The DFT
calculations now performed show in detail the approach of
the electrophilic Ru-complex to the substrate and confirm the
observed enantioface differentiation. Amides and pyridones
with a suitable olefinic double bond were tested as substrates in
the epoxidation reactions, providing significant enantioselectiv-
ities (70−87% ee) and confirming the stereospecifity (syn
addition) of the epoxidation reaction. Full details of our work
are presented in the following sections of this paper.
Figure 1. Chiral sensitizer 1 and model for a chiral bifunctional ligand
I bound to a catalytically active transition metal M (M = metal center,
gray = ligand).
provide moderate to good enantioselectivity in a photoinduced
electron transfer (PET) reaction.19 A related sensitizer, which is
also based on the rigid 1,5,7-trimethyl-3-azabicyclo[3.3.1]-
nonan-2-one skeleton, delivered high enantioselectivities in the
intramolecular [2 + 2]-photocycloaddition of certain 4-alkenyl
substituted quinolones.20 Since the bicyclic lactam skeleton has
been shown to bind to a variety of amides and lactams,18,21 it
was conceived that chiral bifunctional ligands of type I could
prove to be useful in enantioselective transition metal catalysis.
Despite the similarity in design, the mode of action of a
thermal metal-catalyzed vs a photochemical sensitizer-catalyzed
reaction is distinctly different. In the former case the orientation
and distance between the reacting substrate and the transition
metal are crucial for regio- and enantiocontrol, while in the
latter case the sensitizing unit serves only as a steric shield and
hardly influences the regioselectivity of the reaction. In order to
provide high flexibility regarding potential metals and regarding
potential binding modes, a modular approach for the
construction of compounds I (Figure 1) was developed, in
which the metal binding part (in gray) could be installed by
Sonogashira cross-coupling22 or [3 + 2]-cycloaddition23 to an
alkyne. Compounds with the general structure II (Figure 2) are
RESULTS AND DISCUSSION
■
Enantioselective Epoxidation of 3-Alkenylquinolones.
Previous epoxidation studies had only been performed with
literature known 3-vinylquinolone (7a,35 Figure 4) as the
Figure 2. Alkyne templates with a lactam binding site in a general
representation II and in more specific examples 2 and 3.
represented by the specific lactams 224 with a 1,5,7-trimethyl-3-
azabicyclo[3.3.1]nonan-2-one skeleton25 and 326 with an
octahydro-1H-4,7-methanoisoindol-1-one skeleton.27 Both
compounds are available in enantiomerically pure form based
on known procedures either from Kemp’s triacid28 or from the
Diels−Alder product of 6,6-dimethylfulvene and maleic
anhydride.29
Figure 4. Structure of 3-alkenylquinolones 7 employed in this study.
substrate.26 In order to further evaluate the reaction scope
several other substituted 3-alkenylquinolones (Figure 4) were
prepared. A substitution at position C6 or C7 seemed useful to
investigate the electronic influence on the para-positioned
nitrogen atom (C6) or the para-positioned α,β-unsaturated
double bond (C7).
All quinolones 7, with the exception of 7a, have not yet been
reported in the literature. Two routes were employed to access
this compound class. If the corresponding 3-formylquinolone
was known, it was attempted to perform the olefination by a
In preliminary work26 it was shown that the ruthenium
porphyrin complex 4 (Figure 3) can be easily prepared by the
Sonogashira cross-coupling reaction22,30 of alkyne 3 with the
respective monobromo-substituted ruthenium porphyrin,31
which was in turn prepared32 from the known monobromo-
porphyrin33 (see Supporting Information). In addition, the N-
methylated compound 5 and an achiral analogue 6 of
12870
dx.doi.org/10.1021/ja305890c | J. Am. Chem. Soc. 2012, 134, 12869−12878

Journal of the American Chemical Society
Article
Wittig reaction.36 Unsubstituted 3-formylquinolone (8),37 for
example, could be converted into the E-configured olefin 7b by
treatment with the respective phosphorus ylide (Scheme 1).
Most substrates reacted cleanly in good yields with high
enantiomeric excess (85−98% ee). A modification of the R
substituent at the vinylic double bond from hydrogen (entry 1)
to alkyl (entries 2−4) resulted in slighly lower enantiomeric
excesses (86−88% ee). The isopropyl-substituted substrate 7d
(entry 4) exhibited a diminished chemoselectivity despite the
fact that conversion was complete after 24 h. Substitution at
carbon atom C6 turned out to be beneficial for the
enantioselectivity (entries 5−7), and substrates 7e−7g gave
the highest enantioselectivity (95−98% ee) of all the
quinolones. Ester 7f (entry 6) and ester 7i (entry 9) showed
only a reduced solubility in benzene, and they were
consequently employed in a concentration of 10 mM. Still,
the reactions remained incomplete, and unreacted starting
material was recovered after 24 h. Methoxy substitution at
carbon atom C7 (entry 8) was detrimental for the reactivity and
for the product stability. In line with the insufficient
chemoselectivity observed in its epoxidation by mCPBA (vide
supra), the Ru-catalyzed epoxidation of substrate 7h was
sluggish and the resulting epoxide could not be isolated in pure
form. The reaction did not go to completion, and in the crude
product mixture the substrate was detected in significant
amounts (product/substrate = 64/36 by 1H NMR integration)
even after a reaction time of 48 h. The other substituents R2
(entries 8−11) had, apart from the reduced solubility of 7i,
little influence on the reactivity when compared to the parent 3-
vinylquinolone (7a). The corresponding epoxides 10i−k were
produced in good yields and with high enantioselectivity (85−
92% ee). All yields were diminished by the instability of the
epoxides on silica gel or on alumina. Rearrangement products
were observed in some cases, explaining in part why the yields
of the isolated product generally did not exceed 80% despite
the fact that the epoxidation reactions proceeded cleanly and
without a significant byproduct. Compounds of type 10 have
previously not been accessible in an enantioselective fashion.40
The absolute product configuration of epoxides 10 was
assigned based on the previously established configuration of
epoxide 10a.26 The assignment was confirmed by further
studies (vide infra) and by the fact that all epoxides 10 were
consistently levorotatory. The reaction proceeded diastereose-
lectively for E-configured olefins 7b−7d, and only the trans-
configured epoxides 10b−10d were isolated. Since the Z-
configured olefins could not be obtained in pure form, it was
not possible to prove the stereospecifity of the epoxidation for
3-alkenylquinolones 7. However, it was shown that Z-
configured olefins do undergo the enantioselective epoxidation.
The tricyclic substrate 13 for example was readily available by
ring closing metathesis41 from 4-(3-butenyl)-3-vinylquinolone
(12), which in turn was accessible in three steps from ortho-
aminobenzonitrile (11). Due to its insufficient solubility in
benzene, substrate 13 was epoxidized in dichloromethane
(Scheme 3). The reaction proceeded slowly at a substrate
concentration of 5 mM, but it delivered epoxide 14 with an
excellent enantioselectivity of 92% ee. The reaction was
complete after 48 h, and the product was isolated in 62% yield.
The crucial influence of the hydrogen bonding between the
substrate and catalyst 4 was demonstrated by a competition
experiment with quinolone 7a and its N-methylated derivative
15 (Scheme 4). After 4 h the product ratio 10a/16 was 91/9,
and after 24 h it changed slightly to 84/16. The conversion was
almost complete (93% conversion) for substrate 7a after 24 h,
and product 10a was isolated in 74% yield. In an independent
experiment, the oxidation of substrate 15 was performed under
Scheme 1
The Z-olefin was observed as a byproduct but could not be
isolated in pure form. Attempts to obtain the Z-isomer of
compound 7b by Suzuki cross-coupling of 3-bromoquinolone
were more successful. The compound isomerized readily,
however, even in the solid state, which prevented its use in the
epoxidation reaction. The 3-alkenylquinolones 7c, 7g, and 7j
were also obtained by Wittig reaction of the respective
aldehydes (for more details, see the Supporting Information).
An alternative approach was chosen if the corresponding
ortho-amino- or ortho-nitrobenzaldehydes were known or
commercially available. In this case, N-acylation, for the nitro
compounds after a preceding reduction, with the acid chloride
of a β,γ-unsaturated acid, led to an immediate precursor for an
intramolecular aldol condensation, which was performed with
KOH in ethanol.35 As an example the reaction of ortho-
aminobenzaldehyde 938 with 3-butenoyl chloride is shown in
Scheme 1, which led to 3-vinylquinolone 7e in a two-step
sequence. In an analogous fashion the quinolones 7d, 7f, 7h, 7i,
and 7k were synthesized (for more details, see the Supporting
Information).
Quinolones 7 could be oxidized to the respective racemic
epoxides rac-10 with meta-chloroperbenzoic acid (mCPBA) in
dichloromethane. All products could be readily isolated except
for the epoxide derived from quinolone 7h, which was unstable
and underwent unspecific ring-opening reactions. The
enantioselective epoxidation was initially performed with 1
mol % of catalyst 4, but it was found that in most cases 0.2 mol
% of the catalyst was sufficient to obtain optimal results
(Scheme 2). 2,6-Dichloropyridine-N-oxide was used as a
stoichiometric oxidant,39 which was, together with the substrate
(c = 20 mM), dissolved in benzene at ambient temperature.
The catalyst was added as a benzene solution, and the reaction
mixture was stirred for 24 h before workup. The results of these
experiments are summarized in Table 1.
Scheme 2
12871
dx.doi.org/10.1021/ja305890c | J. Am. Chem. Soc. 2012, 134, 12869−12878

Journal of the American Chemical Society
Article
Table 1. Reaction Conditions, Yields, and Enantioselectivities in the Ru-Catalyzed Epoxidation of Quinolones 7 (see Scheme 2)
a
b
c
entry
substrate
R
R1
R2
product
yield [%]
ee [%]
1
7a
7b
7c
7d
7e
H
H
H
H
H
H
H
H
H
H
H
H
10a
10b
10c
10d
10e
10f
72
74
83
60
69
96
2
Me
Et
iPr
H
86
3
86
4
87
5
MeO
MeO2C
Me
H
98
d
e
6
7f
H
49(71)
96
7
7g
7h
H
10g
10h
10i
70
95
f
8
H
MeO
MeO2C
Me
―
45(70)
―
85
d
e
9
7i
H
H
10
11
7j
H
H
10j
65
71
92
7k
H
H
F
10k
92
a
The reactions were conducted at a substrate concentration of 20 mM at ambient temperature in benzene as the solvent, employing 1.1 equiv of 2,6-
b
c
dichloropyridine-N-oxide as the oxidant and 0.2 mol % catalyst 4. The reaction time was 24 h. Yield of isolated product. The enantiomeric excess
was determined by chiral HPLC (see Supporting Information). Due to the low solubility of the substrate the reaction was performed at a substrate
concentration of 10 mM. The reaction remained incomplete after 24 h. The yield in brackets is based on recovered starting material. The product
d
e
f
was formed sluggishly and could not be isolated (see narrative for more information).
and for catalytic olefin epoxidation under aerobic conditions.42
Scheme 3
The use of chiral ruthenium porphyrins for enantioselective
epoxidation reactions began to be explored in the 1990s,43 and
the enantioselective epoxidation with chiral metalloporphyrin
catalysts has remained an active research area to date.44
In most racemic and enantioselective Ru-catalyzed epox-
idation reactions a dioxo ruthenium(VI) intermediate has been
assumed to be the catalytically active intermediate, from which
the oxygen transfer to the olefinic double bond occurs.44 The
other oxo substituent serves as a spectator ligand, which
remains attached to the ruthenium metal in the monooxo
ruthenium(IV) complex, which is subsequently reoxidized to
the dioxo species. In our case, the corresponding dioxo complex
17 was expected to deliver the oxygen atom to the epoxide
within a hydrogen-bound complex 17·7a (Figure 5). Enantio-
face differentiation would be possible if the rotation around the
Scheme 4
carbon−carbon bond, which connects the vinyl group to the
quinolone, was restricted.
the typical conditions for quinolone oxidation with 0.2 mol %
catalyst. It was found that the reaction was close to completion
after 24 h and yielded product rac-16 in 62% yield without any
Figure 5. Preliminary model of the oxygen transfer from the reactive
Ru-complex 17 derived from complex 4 to 3-vinylquinolone (7a) (the
porphyrin skeleton is schematically drawn in gray).
detectable ee (≤5% ee). Similarly, the N-methylated derivative 5
of catalyst 4 (Figure 3) induced only a low enantioselectivity
(14% ee) in the epoxidation of 3-vinylquinolone (7a).26
As illustrated by Scheme 5, an s-cis-configured 3-
DFT Calculations and Mode of Action. Although the
alkenylquinolone would require an Si face attack (relative to
studies with the N-methylated derivatives just described clearly
indicate that there is a two-point hydrogen bond interaction
Scheme 5
between the lactam part of catalyst 4 and the quinolone amide
function at nitrogen atom N1 and carbon atom C2, the exact
mode of the oxygen transfer from the ruthenium porphyrin
complex and the enantioface differentiation was further
investigated. Seminal studies by Groves et al. in the 1980s
had been performed with dioxo(tetramesitylporphyrinato)-
ruthenium(VI), which was shown to be competent for
stoichiometric olefin epoxidation under anaerobic conditions
12872
dx.doi.org/10.1021/ja305890c | J. Am. Chem. Soc. 2012, 134, 12869−12878

Journal of the American Chemical Society
Article
the internal prostereogenic carbon atom of the double bond) of
the oxygen atom to account for the enantioselective formation
of products 10, while in the complex depicted in Figure 5, the
Re face of the s-trans-conformer was to be attacked. In order to
confirm that there is a significant difference in the transition
states leading from quinolones 7 to products 10, theoretical
calculations were performed, in which 3-vinylquinolone (7a)
served as a representative olefin.
For the DFT investigations, the M06L45 density functional
was used in combination with the Stuttgart/Dresden46 basis set
(including the corresponding pseudopotential) for ruthenium
and the 6-31G(d)47 basis set for all other atoms. All calculations
were performed on the untruncated complexes, using
Gaussian0948 with the grid=ultrafine option for integral
evaluation. Since the initial attack of the Ru-bound oxo
substituent on the vinyl group of the substrate is clearly the
selectivity-determining step, we focused our calculations on the
transition states of this process. The nature of these transition
states was verified by analytic computation of vibrational
frequencies: all structures reported below show one imaginary
frequency corresponding to the above-mentioned attack of an
oxo substituent on the quinolone substrate.
Figure 7. Transition state TScis for the attack of an oxo substituent of
17 on the vinyl group of quinolone 7a in an s-cis conformation. The
substrate is bound to the Ru complex by a two-point hydrogen bond.
For better comprehensibility, various parts of the Ru complex are
shown with different visualization styles, although all parts of the
complex have been treated equivalently.
Calculations were initially performed on both the singlet and
the triplet energy surface. No spin contamination was found for
the (unrestricted) singlet structures, validating the treatment of
these systems with the single-reference DFT methodology. The
triplet transition states of the dioxo complex 17 with both 7a
(s-cis) and 7a (s-trans) turned out to be approximately 58 kJ/
mol higher in free energy than their respective singlet
equivalents and were as a consequence not considered to be
relevant in the context of this investigation.
1.85 Å for the favorable TStrans and 1.80 Å for the less favorable
TScis transition state. The Ru−O bond on the site of the attack
is somewhat elongated, with a bond length of 1.83 Å for TStrans
and 1.84 Å for TScis, compared to 1.75 Å for the Ru−O bond
on the site opposite to the attack. The same is true for the C
C bond length of the vinyl subunit, which equals 1.39 Å for
TStrans and 1.40 Å for TScis (compared to 1.34 Å for the isolated
substrate in both the s-trans and s-cis conformation). In both
transition states, there is only a very weak, if any, interaction of
the oxo substituent with the inner carbon atom of the vinyl
group. The respective O−C distances are 2.61 Å for TStrans and
2.53 Å for TScis. As expected, the terminal carbon atom is partly
pyramidalized, while the inner carbon atom remains planar.
The most obvious structural differences between the two
transition states concern the binding of the substrate to the
ligand and, to a lesser extent, the deformation of the C−C triple
bond that links the porphyrin core to the chiral hydrogen bond
template of the ligand. In the more favorable TStrans, the plane
of the quinolone ring is essentially parallel to the plane of the
porphyrin ring. In order to accomplish this, the substrate does
not occupy the plane formed by the hydrogen bond motif of
the chiral ligand, but is bent toward the porphyrin with an angle
of approximately 40°. The CC triple bond of the ligand
shows some deviation from a linear arrangement, with CC
C and CCCporphyrin angles of 172° and 175°, respectively.
As a consequence, the torsion angle of the vinyl group of the
substrate with respect to the quinolone ring is only 6° in the
transition state.
In the less favorable TScis, the quinolone ring is not parallel to
the porphyrin ring or to the hydrogen bond binding motif of
the ligand, but features a slight rotation around an axis
equivalent to the hydrogen bonding direction. The CC triple
bond of the ligand is markedly more linear (with CCC
and CCCporphyrin angles of 178° and 179°, respectively).
The torsion of the vinyl group of the substrate with respect to
the quinolone ring is increased in this case and equals 17°. The
different torsion angles of the vinyl group in the two transition
states seem to indicate that more energy is required to distort
the substrate from the ground state geometry to the geometry
found in the transition state in the case of the less favorable
On the singlet energy surface, the transition state of 17·7a
with the substrate in the s-trans conformation, TStrans (Figure
6), was found to be more stable than the transition state with
Figure 6. Transition state TStrans for the attack of an oxo substituent of
17 on the vinyl group of quinolone 7a in an s-trans conformation. The
substrate is bound to the Ru complex by a two-point hydrogen bond.
For better comprehensibility, various parts of the Ru complex are
shown with different visualization styles, although all parts of the
complex have been treated equivalently.
the substrate in the s-cis conformation, TScis (Figure 7), by 21
kJ/mol in energy (or 12 kJ/mol in free energy). This difference
in energy thus provides a rationalization of the experimental
results mentioned above, as the more favorable transition state
features an Re face attack on the s-trans conformer of the
substrate, leading to the experimentally found enantioselectiv-
ity.
In both cases, the oxo substituent attacks the terminal carbon
atom of the vinyl group. The corresponding O−C distances are
12873
dx.doi.org/10.1021/ja305890c | J. Am. Chem. Soc. 2012, 134, 12869−12878

Journal of the American Chemical Society
Article
transition state. This might explain at least part of the energetic
difference between both cases. In order to verify this
assumption computationally, we calculated the respective
energies required to distort 7a from its equilibrium geometry
to the geometry of the isolated substrate in the transition state
for both the s-trans and the s-cis conformer.49 We found,
however, little difference in these deformation energies (34 kJ/
mol for s-trans and 37 kJ/mol for s-cis). While this type of
analysis certainly has only orientating character, it seems to
indicate that the stronger deformation of the substrate in the
case of TScis is not the deciding factor for the energetic
differences in both transition states. The same analysis for the
ruthenium porphyrin complex, however, yielded deformation
energies of 26 kJ/mol for TStrans and 41 kJ/mol for TScis, and
thus a much larger difference between the two terms. The
distortion of the ruthenium complex consequently seems to be
one important component, which is responsible for the energy
difference of the two transition states.
A second component, which contributes to the energetic
difference between TStrans and TScis, is due to a disparity
between the binding strengths of the two-point hydrogen bond
motifs. While the exact interaction energy is somewhat difficult
to assess considering the rather moderate basis sets employed,
we observed a difference in the interaction distances for both
structures. Thus, while the NHligand−Osubstrate distance is
essentially the same for both cases (1.86 Å for TStrans and
1.85 Å for TScis), the Oligand−HNsubstrate distance is somewhat
longer for TScis (1.81 vs 1.77 Å), indicating a weaker hydrogen
bond. It seems that, for TScis, the slight rotation of the
quinolone ring around the hydrogen-bonding axis, while
necessary to bring the s-cis vinyl group close enough to the
oxo substituent, also reduces the binding strength of the
substrate to the ligand. All issues considered, less deformation
of the binding partners is needed to reach TStrans, and this,
together with the (presumably) stronger hydrogen-bond
interaction energy, serves to explain why this transition state
is more stable and why, experimentally, the enantiomer
corresponding to an Re attack on the s-trans substrate is
found as the product.
It was probed computationally whether any influence of the
quinolone substitution pattern can be detected. More
specifically, the transtition states for the reactions of substrates
7f and 7i were calculated (for details, see Supporting
Information). Most importantly, for both 7f and 7i, again a
preference for TStrans was found by the DFT calculations. The
energetic difference between TStrans and TScis was almost
identical to the one found for 7a in the case of substrate 7f
(approximately 12 kJ/mol) and slightly higher for substrate 7i
(approximately 16 kJ/mol). The large size of the investigated
system prevented the use of more accurate computational
methods. The calculated energetic differences are not
considered sufficiently precise to account for the minor
differences in the experimentally observed enantioselectivities.
Mechanistic Discussion. Based on previous reports about
Ru-catalyzed oxidation reactions,39,43,44 on our own experience
in the field of hydrogen bond mediated enantioselective
reactions,18−21 and on the calculations discussed in the
preceding paragraph, a mechanistic pathway can be proposed,
which is depicted in Scheme 6 with quinolone 7a as a
representative substrate.
Scheme 6
leads to formation of the binary complex 17·7a (Figure 5), in
which oxygen transfer via transition state TStrans (Figure 6) is
rapid. Based on association21b and kinetic20b data obtained for
related hydrogen bound complexes, the rate for the dissociation
of complex 17·7a can be estimated to be on the order of 107 s−1
at ambient temperature. The question of whether oxygen
transfer is faster than dissociation, i.e. whether complex
formation is reversible, cannot be answered at this point in
time. In any case, the transition state TStrans, which leads to the
oxidized product in complex 18·10a, is energetically favored
over any transition state, which operates without hydrogen
bonding interactions (see Scheme 4). Formation of the
enantiomeric product ent-10a could be explained by population
of TScis, which is higher in energy. However, competing
transition states in the selectivity-determining step, which
proceed via no or single-point hydrogen bonding, are also
accessible. For the reaction of N-vinylquinolones 7a, 7e−7g,
7i−7k it appears as if the binding properties are responsible for
the different enantioselectivities observed; i.e. unselective
reactions proceed via competing transition states without a
clear directivity. Indeed, the calculations show no significant
energy differences for the two doubly hydrogen-bound
transitions states irrespective of methoxycarbonyl substitution
in the 6- or 7-position (vide supra). Contrary to that, the
binding properties of N-alkenylquinolones 7a−7d should be
identical and it is likely that formation of the minor
enantiomers occurs in these cases via diastereomorphic
transition states related to the s-cis rotamers of the respective
substrate (Scheme 5). Subsequent dissociation of the product
enables reoxidation of ruthenium(IV) complex 18 to the
catalytically active species 17.
The instability of the product epoxides 10 made quantitative
kinetic analysis of the epoxidation reaction difficult. In addition,
an induction period was observed in the reactions, presumably
due to the fact that compound 4 is initially oxidized to the
catalytically active species. This process is not necessarily
quantitative. In other words, a direct correlation between the
amount of compound 4 and the formation of the putative
catalytically active species 17 does not exist. Lowering the
amount of catalyst 4 to 0.1 mol % under otherwise identical
conditions still led to an almost complete conversion (94%)
after 24 h, and product 10a was isolated in 67% yield. At a
catalyst concentration of 0.05 and 0.01 mol %, the reaction
remained incomplete. Product 10a was isolated together with
the starting material. Product yields were 65% for 0.05 mol %
and 47% for 0.01 mol %. The enantiomeric excess of the
A ruthenium(VI) complex 17 is suggested as the reactive
species, which is formed from compound 4 by oxidation with
2,6-dichloropyridine-N-oxide. An approach of the substrate
12874
dx.doi.org/10.1021/ja305890c | J. Am. Chem. Soc. 2012, 134, 12869−12878

Journal of the American Chemical Society
Article
product remained, within the limits of error ( 1% ee),
unchanged at a catalyst concentration of ≥0.05 mol % showing
the high stereochemical fidelity of the catalyst even at low
concentration. Only at very low catalyst loading (0.01 mol %)
did the enantioselectivity decrease to 92% ee. Turnover
numbers are high. Given that 0.01 mol % of the catalyst
facilitates a substrate conversion of close to 50% after 24 h, the
turnover number exceeds 4000.
Enantioselective Epoxidation of Other Substrates.
From the picture of the preferred transition state TStrans of
the epoxidation reaction, it is evident that any substrate in
which the double bond in the 3-position of the quinolone
cannot readily adopt an s-trans conformation will be difficult to
oxidize. Indeed, the previously mentioned compound 12
(Figure 8), which bears a butenyl group at C4, failed to react
Additional qualitative data were obtained from reactions,
which were interrupted after a reaction time of 4 h (Table 2).
Table 2. Influence of Different Conditions on the Ru-
Catalyzed Epoxidation of 3-Vinylquinolone (7a) after 4 h
a
b
c
d
e
entry catalyst
equiv oxid.
additive 10a/7a yield [%] ee [%]
Figure 8. Structures of 3-vinyl-substituted quinolones 12, 19, and
pyridone 20a.
1
2
3
4
5
4
4
4
4
6
1.1
0.55
2.2
1.1
1.1

50/50
23/77
79/21
53/47
19/81
37
14
57
39
15
96
93
96
93



cleanly under the typical reaction conditions, which were
applied to substrates 7. In addition, the reaction was by far less
enantioselective. The respective epoxide was obtained in a yield
of 39% after a reaction time of 24 h and with 31% ee.
f
3-EQ

a
The reactions were conducted at a substrate concentration of 20 mM
at ambient temperature in benzene as the solvent, employing 2,6-
The highly ordererd transition state of the epoxidation
reaction also suggests that the epoxidation occurs with high
regioselectivity if a substrate contains two electronically similar
vinylic double bonds. Compound 19 for example was shown to
undergo the epoxidation with catalyst 4 in a regioselectivity of
91/9 in favor of the double bond in the 3-position (88% ee).26
Under the same reaction conditions, the use of catalyst 6 led to
a sluggish conversion, which was not complete even after 40 h
and which resulted in a regioisomeric ratio of 62/38.
Given the lactam binding mode operating in the case of
quinolones 7, it was expected that the analogous 3-
alkenylpyridones would react with similar enantioselectivity as
the respective quinolones. It must be noted, however, that
epoxides of 3-alkenylpyridones have to the best of our
knowledge not yet been described in the literature. It was
consequently not unexpected that the parent 3-vinylpyridone
(20a)50 delivered an epoxide, which was impossible to isolate
by column chromatography on silica or on alumina. In addition,
like in the case of substrate 7h, the reaction proceeded slowly
and with low chemoselectivity. The 5-substituted pyridones
20b51 and 20c52 reacted more efficiently, but the isolation
problems persisted. Despite the fact that the reaction was
complete after 24 h, the epoxides could not be purified and
determination of the ee was not possible. Eventually the
intermediate epoxides were opened with a sulfur nucleophile
(para-thiocresol, ArSH),53 and the enantiomeric purity of the
resulting secondary alcohols was determined (Scheme 7). The
enantioselectivity was high in both cases, and the analogy to the
quinolone examples seems to hold.
dichloropyridine-N-oxide as the oxidant and 0.2 mol % catalyst. The
b
reaction time was 4 h. Equivalents of 2,6-dichloropyridine-N-oxide
c
employed in the reaction. The ratio of product/substrate was
d
e
1
determined by H NMR integration. Yield of isolated product. The
f
enantiomeric excess was determined by chiral HPLC. The reaction
was performed in the presence of stoichiometric amounts (1 equiv) of
3-ethylquinolone (3-EQ).
The initial reaction rate depends significantly on the amount of
oxidant being used in the process (entries 1−3). After 4 h and
with 1.1 equiv of 2,6-dichloropyridine-N-oxide the conversion
of substrate 7a was complete to a degree of roughly 50%
(product/substrate = 50/50) in the presence of 0.2 mol % of
catalyst 4. With 0.55 equiv of the oxidant the conversion
dropped to 23% (entry 2) and it increased to 79% with 2.2
equiv of the oxidant (entry 3). The data suggest that the
oxidant is involved in the rate-determining step, which appears
to be the oxidation of ruthenium(IV) complex 18 to the
ruthenium(VI) complex 17.
The fact that the epoxidation step (17·7a → 18·10a) is not
rate determining is further supported by the fact that the
catalytic performance was not influenced by the addition of
another 3-substituted quinolone, which competes with the
binding site. 3-Ethylquinolone (3-EQ) was added in stoichio-
metric amounts to the catalytic epoxidation 7a → 10a, and the
reaction was stopped after 4 h (entry 4). The product/substrate
ratio was identical (product/substrate = 53/47) as compared to
the reaction without an additive (entry 1). The enantiomeric
excess was slightly lower (93% ee). The rate but not the
enantioselectivity was influenced at higher 3-EQ concentra-
tions. After 10 h and with 5 equiv of 3-EQ (see Supporting
Information), the product/substrate ratio decreased from 43/
57 (no additive) to 21/79 and, with 10 equiv, to 15/85. There
was only a minor deviation in the enantiomeric excess (92%
and 91% ee). These data suggest that reoxidation can only
occur when the binding site is vacant. Blocking the binding site
retards the reoxidation of 18 to 17. Regarding the reoxidation
from Ru(IV) to Ru(VI), the situation for catalyst 6 is likely to
be different because the ruthenium center is much easier to
approach by the oxidant. The lower conversion in this case
(entry 5) can be explained by a slower epoxidation reaction as
compared to the rapid epoxidation within complex 17·7a.
Scheme 7
12875
dx.doi.org/10.1021/ja305890c | J. Am. Chem. Soc. 2012, 134, 12869−12878

Journal of the American Chemical Society
Article
While both quinolones and pyridones display a lactam
binding site, it seemed appropriate to investigate whether
primary alkenoic amides would also be amenable to an
enantioselective epoxidation reaction. Literature known amide
2254 was prepared and subjected to the standard reaction
conditions (Table 1) employing, however, 1.5 mol% of catalyst
4. The higher catalyst loading was required to enforce a
complete reaction after three days. The reaction yielded 61% of
cis-epoxide 23 in 70% ee (Scheme 8). For this example the
Proof of the absolute configuration was obtained for
compound 26 again by hydrogenolysis to a secondary alcohol.
The resulting product 27 could be obtained also from (S)-
methyl phenyllactate (see Supporting Information). In this
case, however, the synthetic product showed the opposite
absolute configuration and the (R)-configuration was con-
sequently assigned to the stereogenic center in compound 27.
Apparently carbamate 25 prefers a stretched conformation in
the reactive complex (Figure 9) avoiding 1,3-allylic strain
between the phenyl group and the oxygen atom.
Scheme 8
Figure 9. Model for the oxygen transfer to the olefinic double bond in
primary amide 25 (the porphyrin skeleton is schematically drawn in
gray).
stereospecifity of the epoxidation process could be proven
because the trans-alkenoic amide was also synthetically
accessible.55 It has been previously observed that trans-
substituted styrenes are less reactive in Ru-catalyzed
epoxidation reactions.39b In line with this observation,
conversion of the trans-substrate was very slow (16%
conversion after four days), and product formation remained
incomplete. However, the only detectable product was the
trans-epoxide.
The last two results show that primary amides can be used in
enantioselective reactions, in which the enantioselectivity is
transferred from a rigid chiral backbone to the reaction center
via hydrogen bonds. However, the higher conformational
flexibility of these substrates is detrimental to the required
enantioface differentiation.
The absolute configuration of epoxide 23 was established
upon hydrogenolysis to the secondary alcohol 24, which could
be independently obtained from enantiomerically pure (S)-
methyl phenyllactate (see Supporting Information). The
stereochemical identity of the compounds was proven by
chiral HPLC and chiroptical methods (specific rotation). The
outcome of this reaction supports the model for oxygen transfer
in the hydrogen-bound complex (Figure 5). Amide 22 is able to
adopt a conformation similar to quinolones 7 and exposes the
double bond to the ruthenium(VI) dioxo complex 17 in the
same fashion as the respective s-trans conformers of 7. The
lower selectivity is likely due to the higher degree of rotational
freedom. While there is only one single bond in quinolones 7,
around which a rotation is possible, amide 22 contains two
single bonds between the amide group and the reactive double
bond. The situation becomes even more complicated for
carbamate 2556 (Scheme 9), in which there are three single
bonds between the binding motif and the reacting center. In
this case, the enantioselectivity of the epoxidation further
decreased and product 26 was obtained in a good yield of 69%
but with only moderate enantioselectivity (45% ee).
CONCLUSION
■
The present study provides convincing evidence that the
presence of two hydrogen bonding sites at a chiral bifunctional
ligand allows for highly enantioselective reactions, which occur
at a spatially remote metal center. Substrate coordination via
hydrogen bonding lowers the transition state of the metal
catalyzed process so that it occurs almost exclusively within the
hydrogen-bound complex. The chiral information is trans-
mitted via the chiral ligand backbone over a distance of more
than 7.0 Å. The general binding mode enables a substrate
variation as long as the reactive center is exposed to the
catalytically active metal in a defined fashion. Although the
current study is the first of its kind and has limited predictive
character, further optimization should lead to supramolecular
catalysts, which are able to attack less activated reactive centers
than double bonds. In addition it should be possible to change
the regioselectivity of attack by shortening or lengthening the
distance between the hydrogen bonding site and the catalyti-
cally active center. Work along these lines continues in our
laboratories.
ASSOCIATED CONTENT
* Supporting Information
Detailed experimental procedures, characterization data for new
compounds, and details of the DFT calculations. This material
is available free of charge via the Internet at http://pubs.acs.org.
■
Scheme 9
S
AUTHOR INFORMATION
Corresponding Author
thorsten.bach@ch.tum.de
■
Notes
The authors declare no competing financial interest.
12876
dx.doi.org/10.1021/ja305890c | J. Am. Chem. Soc. 2012, 134, 12869−12878

Journal of the American Chemical Society
Article
(15) Agarkov, A.; Greenfield, S.; Xie, D.; Pawlick, R.; Starkey, G.;
Gilbertson, S. R. Biopolymers (Pept. Sci.) 2006, 84, 48−73.
(16) (a) Breuil, P.-A. R.; Patureau, F. W.; Reek, J. N. H. Angew.
Chem., Int. Ed. 2009, 48, 2162−2165. (b) Breuil, P.-A. R.; Reek, J. N.
H. Eur. J. Org. Chem. 2009, 6225−6230.
(17) Dydio, P.; Rubay, C.; Gadzikwa, T.; Lutz, M.; Reek, J. N. H. J.
Am. Chem. Soc. 2011, 133, 17176−17179.
ACKNOWLEDGMENTS
■
This project was supported by the Deutsche Forschungsgemein-
schaft (Ba 1372-17) and by the TUM Graduate School. We
thank Wacker-Chemie (Munich) for the donation of chemicals.
S.M.H. thanks the Fonds der Chemischen Industrie for a Liebig
fellowship.
(18) Review: Breitenlechner, S.; Selig, R.; Bach, T. In Organocatalysis:
Ernst Schering Foundation Symposium Proceedings; Reetz, M. T., List, B.,
Jaroch, S., Weinmann, H., Eds.; Springer: Heidelberg, 2008; pp 255−
279.
REFERENCES
■
(1) Hartwig, J. F. Organotransition Metal Chemistry: From Bonding to
Catalysis; University Science: Sausalito, CA, 2010.
(2) van Leeuwen, P. W. N. M. Supramolecular Catalysis; Wiley-VCH:
(19) Bauer, A.; Westkamper, F.; Grimme, S.; Bach, T. Nature 2005,
̈
436, 1139−1140.
Weinheim, 2008.
(20) (a) Muller, C.; Bauer, A.; Bach, T. Angew. Chem., Int. Ed. 2009,
48, 6640−6642. (b) Muller, C.; Maturi, M. M.; Bauer, A.; Cuquerella,
M. C.; Miranda, M. A.; Bach, T. J. Am. Chem. Soc. 2011, 133, 16689−
̈
(3) Review on enzyme mimics based upon supramolecular
coordination chemistry: Wiester, M. J.; Ulmann, P. A.; Mirkin, C. A.
Angew. Chem., Int. Ed. 2011, 50, 114−137.
(4) For an excellent recent review, see: Carboni, S.; Gennari, C.;
Pignatoro, L.; Piarulli, U. Dalton Trans. 2011, 40, 4355−4373.
(5) For a short review on scaffolding ligands, see: Tan, K. L.; Sun, X.;
Worth, A. D. Synlett 2012, 23, 321−325.
(6) Review: Crabtree, R. H. New J. Chem. 2011, 35, 18−23.
(7) Reviews and monographs on cytochrome P450 mimics:
(a) Mansuy, D. Pure Appl. Chem. 1987, 59, 759−770. (b) Mansuy,
D. Pure Appl. Chem. 1994, 66, 737−744. (c) Feiters, M. C.; Rowan, A.
E.; Nolte, R. J. M. Chem. Soc. Rev. 2000, 29, 375−384. (d) Meunier, B.
Biomimetic Oxidations Catalyzed by Transition Metal Complexes;
Imperial College Press: London, 2000. (e) Groves, J. T. Proc. Natl.
Acad. Sci. U.S.A. 2003, 100, 3569−3574. (f) Ortiz de Montellano, P. R.
Cytochrome P450: Structure, Mechanism, and Biochemistry, 3rd ed.;
Kluwer: New York, 2005.
̈
16697.
(21) (a) Albrecht, D.; Vogt, F.; Bach, T. Chem.Eur. J. 2010, 16,
4284−4296. (b) Bakowski, A.; Dressel, M.; Bauer, A.; Bach, T. Org.
Biomol. Chem. 2011, 9, 3516−3529. (c) Austin, K. A. B.; Herdtweck,
E.; Bach, T. Angew. Chem., Int. Ed. 2011, 50, 8416−8419.
(22) Reviews: (a) Sonogashira, K. J. Organomet. Chem. 2002, 653,
46−49. (b) Negishi, E.; Anastasia, L. Chem. Rev. 2003, 103, 1979−
2017. (c) Chinchilla, R.; Naj
́
era, C. Chem. Rev. 2007, 107, 874−922.
(d) Chinchilla, R.; Najera, C. Chem. Soc. Rev. 2011, 40, 5084−5121.
́
(23) Reviews: (a) Moses, J. E.; Moorhouse, A. D. Chem. Soc. Rev.
2007, 36, 1249−1262. (b) Meldal, M.; Tornøe, C. W. Chem. Rev.
2008, 108, 2952−3015.
(24) (a) Voss, F.; Bach, T. Synlett 2010, 1493−1496. (b) Voss, F.;
Herdtweck, E.; Bach, T. Chem. Commun. 2011, 47, 2137−2139.
(c) Voss, F.; Vogt, F.; Herdtweck, E.; Bach, T. Synthesis 2011, 961−
971.
(8) For artificial metalloenzymes, which act by tethering a transition
metal to a host protein via hydrogen bonds, see: Ward, T. R. Acc.
Chem. Res. 2011, 44, 47−57.
(25) For the use of this skeleton in molecular recognition studies,
see: (a) Rebek, J., Jr.; Askew, B.; Ballester, P.; Buhr, C.; Costero, S.;
Jones, S.; Nemeth, D.; Williams, D. J. Am. Chem. Soc. 1987, 109,
6866−6867. (b) Jeong, K.-S.; Parris, K.; Ballester, P.; Rebek, J., Jr.
Angew. Chem., Int. Ed. 1990, 29, 555−556. (c) Jeong, K. S.; Tjivikua,
T.; Muehldorf, A.; Deslongchamps, G.; Famulok, M.; Rebek, J., Jr. J.
Am. Chem. Soc. 1990, 113, 201−209.
(9) For the use of hydrogen bonds in organocatalysis, see: Taylor, M.
S.; Jacobsen, E. N. Angew. Chem., Int. Ed. 2006, 45, 1520−1543.
(10) For selective oxidation reactions mediated by other noncovalent
interactions except hydrogen bonds, see: (a) Breslow, R.; Zhang, X.;
Huang, Y. J. Am. Chem. Soc. 1997, 119, 4535−4536. (b) Yang, J.;
Breslow, R. Angew. Chem., Int. Ed. 2000, 39, 2692−2694. (c) Yang, J.;
Gabriele, B.; Belvedere, S.; Huang, Y.; Breslow, R. J. Org. Chem. 2002,
67, 5057−5067. (d) Fang, Z; Breslow, R. Org. Lett. 2006, 8, 251−254.
(11) (a) Grotjahn, D. B.; Miranda-Soto, V.; Kragulj, E. J.; Lev, D. A.;
Erdogan, G.; Zeng, X.; Cooksy, A. L. J. Am. Chem. Soc. 2008, 130, 20−
21. (b) Grotjahn, D. B.; Kragulj, E. J.; Zeinalipour-Yazdi, C. D.;
Miranda-Soto, V.; Lev, D. A.; Cooksy, A. L. J. Am. Chem. Soc. 2008,
130, 10860−10861.
(26) Fackler, P.; Berthold, C.; Voss, F.; Bach, T. J. Am. Chem. Soc.
2010, 132, 15911−15913.
(27) For the use of this skeleton in molecular recognition studies,
see: (a) Lonergan, G. D.; Riego, J.; Deslongchamps, G. Tetrahedron
Lett. 1996, 37, 6109−6112. (b) Lonergan, D. G.; Halse, J.;
Deslongchamps, G. Tetrahedron Lett. 1998, 39, 6865−6868. (c) Lone-
rgan, D. G.; Deslongchamps, G. Tetrahedron 1998, 54, 14041−14052.
(28) Kemp, D. S.; Petrakis, K. S. J. Org. Chem. 1981, 46, 5140−5143.
(29) Corwin, L. R.; McDaniel, D. M.; Bushby, R. J.; Berson, J. A. J.
Am. Chem. Soc. 1980, 102, 276−287.
(12) (a) Das, S.; Incarvito, C. D.; Crabtree, R. H.; Brudvig, G. W.
Science 2006, 312, 1941−1943. (b) Das, S.; Brudvig, G. W.; Crabtree,
R. H. J. Am. Chem. Soc. 2008, 130, 1628−1637. (c) Hull, J. F.; Sauer, E.
L. O.; Incarvito, C. D.; Faller, J. W.; Brudvig, G. W.; Crabtree, R. H.
Inorg. Chem. 2009, 48, 488−495.
(30) Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetrahedron Lett. 1975,
16, 4467−4470.
(31) Review on ruthenium porphyrin complexes: Simmoneaux, G.;
Le Maux, P. Coord. Chem. Rev. 2002, 228, 43−60.
(32) For the functionalization of porphyrins, see: (a) Senge, M. O.;
Shaker, Y. M.; Pintea, M.; Ryppa, C.; Hatscher, S. S.; Ryan, A.;
Sergeeva, Y. Eur. J. Org. Chem. 2010, 237−258. (b) Senge, M. O.
Chem. Commun. 2011, 47, 1943−1960.
(33) Liddell, P. A.; Gervaldo, M.; Bridgewater, J. W.; Keirstead, A. E.;
Lin, S.; Moore, T. A.; Moore, A. L.; Gust, D. Chem. Mater. 2008, 20,
135−142.
(34) For pioneering contributions to the enantioselective, transition
metal catalyzed epoxidation of olefins, see: (a) Katsuki, T.; Sharpless,
K. B. J. Am. Chem. Soc. 1980, 102, 5974−5976. (b) Groves, J. T.;
Myers, R. S. J. Am. Chem. Soc. 1983, 105, 5791−5796. (c) Hanson, R.
M.; Sharpless, K. B. J. Org. Chem. 1986, 51, 1922−1925. (d) Zhang,
W.; Loebach, J. L.; Wilson, S. R.; Jacobsen, E. N. J. Am. Chem. Soc.
1990, 112, 2801−2803. (e) Irie, R.; Noda, K.; Ito, Y.; Matsumoto, N.;
Katsuki, T. Tetrahedron Lett. 1990, 31, 7345−7348.
̌
(13) (a) Smejkal, T.; Breit, B. Angew. Chem., Int. Ed. 2008, 47, 3946−
̌
3949. (b) Smejkal, T.; Gribkov, D.; Geier, J.; Keller, M.; Breit, B.
Chem.Eur. J. 2010, 16, 2470−2478. (c) Dydio, P.; Dzik, W. I.; Lutz,
M.; de Bruin, B.; Reek, J. H. N. Angew. Chem., Int. Ed. 2011, 50, 396−
400.
(14) For additional studies on the directing effect of hydrogen bonds
in transition metal catalyzed reactions, see: (a) Coolen, H. K. A. C.;
Meuwis, J. A. M.; van Leeuwen, P. W. N. M.; Nolte, R. J. M. J. Am.
Chem. Soc. 1995, 117, 11906−11913. (b) Jon
́
sson, S. S.; Odille, F. G.
J.; Norrby, P.-O.; Warnmark, K. Chem. Commun. 2005, 549−551.
̈
́
(c) Jonsson, S.; Odille, F. G. J.; Norrby, P.-O.; Warnmark, K. Org.
̈
Biomol. Chem. 2006, 4, 1927−1948. (d) Zhu, S.; Ruppel, J. V.; Lu, H.;
Wojtas, L.; Zhang, X. P. J. Am. Chem. Soc. 2008, 130, 5042−5043.
(e) Lu, Y.; Johnstone, T. C.; Arndtsen, B. A. J. Am. Chem. Soc. 2009,
131, 11284−11285. (f) Dzik, W. I.; Xu, X.; Zhang, X. P.; Reek, J. N.
H.; de Bruin, B. J. Am. Chem. Soc. 2010, 132, 10891−10902 and
references cited therein.
12877
dx.doi.org/10.1021/ja305890c | J. Am. Chem. Soc. 2012, 134, 12869−12878

Journal of the American Chemical Society
Article
(35) Shanmugan, P.; Palaniappan, R. Z. Naturforsch. 1973, 28, 196−
(51) The compound was prepared by Suzuki cross-coupling from the
respective bromide: Novartis A. G. WO2008/58967 A1, 2008 (see
Supporting Information).
199.
(36) (a) Wittig, G.; Schollkopf, U. Chem. Ber. 1954, 87, 1318−1330.
̈
(52) The compound was prepared by Suzuki cross-coupling from the
respective bromide: Wishka, D. G.; Walker, D. P.; Yates, K. M.; Reitz,
S. C.; Jia, S.; Myers, J. K.; Olson, K. L.; Jacobsen, E. J.; Wolfe, M. L.;
Groppi, V. E.; Hanchar, A. J.; Thornburgh, B. A.; Cortes-Burgos, L. A.;
Wong, E. H. F.; Staton, B. A.; Raub, T. J.; Higdon, N. R.; Wall, T. M.;
Hurst, R. S.; Walters, R. R.; Hoffmann, W. E.; Hajos, M.; Franklin, S.;
Carey, G.; Gold, L. H.; Cook, K. K.; Sands, S. B.; Zhao, S. X.; Soglia, J.
R.; Kalgutkar, A. S.; Arneric, S. P.; Rogers, B. N. J. Med. Chem. 2006,
49, 4425−4426 (see Supporting Information)..
(b) Hollywood, F.; Suschitzky, H. Synthesis 1982, 662−665.
(37) El-Sayed, O. A.; Aboul-Enein, H. Y. Arch. Pharm. Pharm. Med.
Chem. 2001, 334, 117−120.
(38) Tsoungas, P. G.; Searcey, M. Tetrahedron Lett. 2001, 42, 6589−
6592.
(39) (a) Higuchi, T.; Ohtake, H.; Hirobe, M. Tetrahedron Lett. 1989,
30, 6545−6548. (b) Ohtake, H.; Higuchi, T.; Hirobe, M. Heterocycles
1995, 40, 867−903.
(40) For the epoxidation of 3-styrylquinolones, which was catalyzed
by a chiral Ru-pybox complex and which proceeded with low
enantioselectivity, see ref 24c.
(41) (a) Schwab, P.; Grubbs, R. H.; Ziller, J. W. J. Am. Chem. Soc.
1996, 118, 100−110. (b) Trnka, T. M.; Grubbs, R. H. Acc. Chem. Res.
2001, 34, 18−29. (c) Arrayas, R. G.; Alcudia, A.; Liebeskind, L. S. Org.
Lett. 2001, 3, 3381−3384.
(42) (a) Groves, J. T.; Quinn, R. J. Am. Chem. Soc. 1985, 107, 5790−
5792. (b) Groves, J. T.; Quinn, R. J. Inorg. Chem. 1984, 23, 3844−
3846.
(53) Node, M.; Nishide, K.; Shigeta, Y.; Obata, K.; Shiraki, H.;
Kunishige, H. Tetrahedron 1997, 53, 12883−12984.
(54) Takami, K.; Yorimitsu, H.; Oshima, K. Org. Lett. 2004, 6, 4555−
4558.
(55) Stoermer, R.; Stockmann, R. Chem. Ber. 1914, 47, 1786−1793.
(56) The compound was prepared from the respective allylalcohol:
Pavlakos, E.; Georgiou, T.; Tofi, M.; Montagnon, T.;
Vassilikogiannakis, G. Org. Lett. 2009, 11, 4556−4559 (see Supporting
Information).
(43) (a) Gross, Z.; Ini, S.; Kapon, M.; Cohen, S. Tetrahedron Lett.
1996, 37, 7325−7328. (b) Gross, Z.; Ini, S. J. Org. Chem. 1997, 62,
5514−5521. (c) Frauenkron, M.; Berkessel, A. J. Chem. Soc., Perkin
Trans. 1 1997, 2265−2266. (d) Lai, T.-S.; Zhang, R.; Cheung, K.-K.;
Kwong, H.-L.; Che, C.-M. Chem. Commun. 1998, 1583−1584. (e) Lai,
T.-S.; Kwong, H.-L.; Zhang, R.; Che, C.-M. J. Chem. Soc., Dalton Trans.
1998, 3559−3564. (f) Ini, S.; Gross, Z. Org. Lett. 1999, 1, 2077−2080.
(g) Zhang, R.; Yu, W.-Y.; Lai, T.-S.; Che, C.-M. Chem. Commun. 1999,
409−410. (h) Zhang, R.; Yu, W.-Y.; Wong, K.-Y.; Che, C.-M. J. Org.
Chem. 2001, 71, 8145−8153. (i) Zhang, J.-L.; Liu, Y.-L.; Che, C.-M.
Chem. Commun. 2002, 2906−2907. (j) Le Maux, P.; Lukas, M.;
Simonneaux, G. J. Mol. Catal. A: Chem. 2003, 206, 95−103.
(k) Berkessel, A.; Kaiser, P.; Lex, J. Chem.Eur. J. 2003, 9, 4746−
4756.
NOTE ADDED AFTER ASAP PUBLICATION
■
The toc/abstract graphic was incorrect in the version published
ASAP July 23, 2012. The correct version reposted August 1,
2012.
(44) Reviews: (a) Meunier, B. Chem. Rev. 1992, 92, 1411−1456.
̀
(b) Rose, E.; Andrioletti, B.; Zrig, S.; Quelquejeu-Etheve, M. Chem.
Soc. Rev. 2005, 34, 573−583. (c) Che, C.-M.; Huang, J.-S. Chem.
Commun. 2009, 3996−4015.
(45) Zhao, Y.; Truhlar, D. G. J. Chem. Phys. 2006, 125, 194101−1−
194101−18.
(46) Dolg, M.; Wedig, U.; Stoll, H.; Preuss, H. J. Chem. Phys. 1987,
86, 866−872.
(47) Hehre, W. J.; Radom, L.; Schleyer, P. v. R.; Pople, J. A. Ab Initio
Molecular Orbital Theory; Wiley: New York, 1986.
(48) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, 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., Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega,
N.; Millam, N. J.; 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 B.01; Gaussian, Inc.: Wallingford, CT, 2009.
(49) For the activation strain model of chemical reactivity, see:
(a) Bickelhaupt, F. M. J. Comput. Chem. 1999, 20, 114−128. (b) van
Zeist, W.-J.; Bickelhaupt, F. M. Org. Biomol. Chem. 2010, 8, 3118−
3127.
(50) Sing, B. K.; Cavalluzzo, C.; Van Der Eycken, E.; Parmar, V. S.;
De Maeyer, M.; Debyser, Z. Eur. J. Org. Chem. 2009, 27, 4589−4592.
12878
dx.doi.org/10.1021/ja305890c | J. Am. Chem. Soc. 2012, 134, 12869−12878