Probing competitive enantioselective approach vectors operating in the Jacobsen-Katsuki epoxidation: A kinetic study of methyl-substituted styrenes

Article abstract of DOI:10.1021/ja051851f

This paper describes a study of reactivity and enantioselectivity for a series of methyl-substituted styrenes in the Jacobsen-Katsuki (Mn(salen)-catalyzed) epoxidation reaction. Competition experiments provided kinetic data for the reactivity of the seven possible methyl-substituted styrenes (mono-, di- and trisubstituted) relative to styrene itself, ee values were measured by chiral GC, and absolute configurations were secured by chemical correlation. Of particular interest was the switch in absolute configuration at the benzylic position of the epoxides derived from (Z)- and (E)-α,β-dimethylstyrene, respectively. The results could be rationalized in terms of an approach vector with the phenyl substituent proximal to the salen. As opposed to alkyl groups, a proximal phenyl group has very little effect on the rate of the reaction. Consideration of distal vs proximal approach allows prediction of absolute stereochemistry as a function of alkene substitution pattern. Trisubstituted alkenes with one phenyl group cis to the alkene hydrogen can be identified as a favored substrate class in the title reaction, with both rate and selectivity close to the classic (Z)-β-substituted styrene substrates.

Full text of DOI:10.1021/ja051851f

Published on Web 09/13/2005
Probing Competitive Enantioselective Approach Vectors
Operating in the Jacobsen-Katsuki Epoxidation: A Kinetic
Study of Methyl-Substituted Styrenes
Peter Fristrup, Brian B. Dideriksen, David Tanner, and Per-Ola Norrby*
Contribution from the Department of Chemistry, Technical UniVersity of Denmark,
Building 201 KemitorVet, DK-2800 Kgs. Lyngby, Denmark
Received March 23, 2005; E-mail: pon@kemi.dtu.dk
Abstract: This paper describes a study of reactivity and enantioselectivity for a series of methyl-substituted
styrenes in the Jacobsen-Katsuki (Mn(salen)-catalyzed) epoxidation reaction. Competition experiments
provided kinetic data for the reactivity of the seven possible methyl-substituted styrenes (mono-, di- and
trisubstituted) relative to styrene itself, ee values were measured by chiral GC, and absolute configurations
were secured by chemical correlation. Of particular interest was the switch in absolute configuration at the
benzylic position of the epoxides derived from (Z)- and (E)-R,â-dimethylstyrene, respectively. The results
could be rationalized in terms of an approach vector with the phenyl substituent proximal to the salen. As
opposed to alkyl groups, a proximal phenyl group has very little effect on the rate of the reaction.
Consideration of distal vs proximal approach allows prediction of absolute stereochemistry as a function of
alkene substitution pattern. Trisubstituted alkenes with one phenyl group cis to the alkene hydrogen can
be identified as a favored substrate class in the title reaction, with both rate and selectivity close to the
classic (Z)-â-substituted styrene substrates.
metal catalyst systems,8 as well as methodology based on chiral
dioxiranes.9 Of these methods, the Jacobsen-Katsuki [Mn-
Introduction
Epoxides have fittingly been described as being “one of the
main muscles” of organic synthesis,1 since a wide range of regio-
and stereoselective ring opening reactions are available2 for the
conversion of epoxides to useful (chiral) intermediates. In
addition, the epoxide ring is an important structural element in
the pharmacophores of certain bioactive natural products, e.g.,
dynemicin A,3 neocarzinostatin,4 and the epothilones.5 Practical
and general methods for the enantioselective epoxidation of
prostereogenic alkenes are thus highly desirable, and the first
major breakthrough came in the early 1980s with the discovery
and development of the Sharpless titanium-catalyzed asymmetric
epoxidation of allylic alcohols.6 This prompted an ongoing
search for other catalyst systems capable of the enantioselective
epoxidation of “nonfunctionalized” alkenes, i.e., substrates
lacking a functional group capable of preorganizing the catalyst.
Significant recent advances toward this challenging goal include
the development of chiral (salen)-metal7 or chiral (porphyrin)-
(salen)]-catalyzed epoxidation reaction in particular has enjoyed
increasing use in target-oriented synthesis.10 As for any catalytic
system, rational development of even better protocols for
enantioselective epoxidation will rely, at least in part, on
information garnered from mechanistic11,12 and theoretical
(7) (a) Zhang, W.; Loebach, J. L.; Wilson, S. R.; Jacobsen, E. N. J. Am. Chem.
Soc. 1990, 112, 2801. (b) Irie, R.; Noda, K.; Ito, Y.; Matsumoto, N.; Katsuki,
T. Tetrahedron Lett. 1990, 31, 7345. (c) Katsuki, T. In Catalytic Asymmetric
Synthesis, 2nd ed.; Ojima, I., Ed.; Wiley-VCH: New York, 2000; Chapter
6B. (d) Jacobsen, E. N.; Wu, M. H. In ComprehensiVe Asymmetric
Catalysis; Jacobsen, E. N., Pfaltz, A., Yamamoto, H., Eds.; Springer-
Verlag: 1999; Chapter 18.2.
(8) (a) Collman, J. P.; Zhang, X. M.; Lee, V. J.; Uffelman, E. S.; Brauman, J.
I. Science 1993, 261, 1404. (b) Groves, J. T.; Viski, P. J. Org. Chem. 1990,
55, 3628.
(9) (a) Shi, Y. Acc. Chem. Res. 2004, 37, 488. (b) Adam, W.; Saha-Mo¨ller, C.
R.; Zhao, C.-G. Org. React. 2002, 61, Chapter 2.
(10) For examples, see (a) Nicolaou, K. C.; Safina, B. S.; Funke, C.; Zak, M.;
Ze´cri, F. J. Angew. Chem., Int. Ed. 2002, 41, 1937. (b) Yoshida, M.; Ismail,
M. A.-H.; Nemoto, H.; Ihara, M. J. Chem. Soc., Perkin Trans. 1 2000,
2629. (c) Chen, Y.; Evarts, J. B., Jr.; Torres, E.; Fuchs, P. L. Org. Lett.
2002, 4, 3571. (d) Coleman, R. S.; Garg, R. Org. Lett. 2001, 3, 3487.
(11) (a) Engelhardt, U.; Linker, T. Chem. Commun. 2005, 1152. (b) Linde, C.;
Koliai, N.; Norrby, P. O.; A° kermark, B. Chem.sEur. J. 2002, 8, 2568. (c)
Adam, W.; Roschmann, K. J.; Saha-Moller, C. R.; Seebach, D. J. Am. Chem.
Soc. 2002, 124, 5068. (d) Adam, W.; Mock-Knoblauch, C.; Saha-Moller,
C. R.; Herderich, M. J. Am. Chem. Soc. 2000, 122, 9685. (e) Palucki, M.;
Finney, N. S.; Pospisil, P. J.; Guler, M. L.; Ishida, T.; Jacobsen, E. N. J.
Am. Chem. Soc. 1998, 120, 948. (f) Linde, C.; Arnold, M.; Norrby, P. O.;
A° kermark, B. Angew. Chem., Int. Ed. Engl. 1997, 36, 1723. (g) Finney, N.
S.; Pospisil, P. J.; Chang, S.; Palucki, M.; Konsler, R. G.; Hansen, K. B.;
Jacobsen, E. N. Angew. Chem., Int. Ed. Engl. 1997, 36, 1720. (h)
Feichtinger, D.; Plattner, D. A. Angew. Chem., Int. Ed. Engl. 1997, 36,
1718. (i) Jacobsen, E. N.; Deng, L.; Furukawa, Y.; Martinez, L. E.
Tetrahedron 1994, 50, 4323.
(1) Seebach, D.; Weidmann, B.; Wilder, L. In Modern Synthetic Methods, 1983;
Scheffold, R., Ed.; Otto Salle Verlag: Frankfurt, 1983; p 323.
(2) For reviews, see: (a) Jacobsen, E. N. Acc. Chem. Res. 2000, 33, 421. (b)
Taylor, S. K. Tetrahedron 2000, 56, 1149. (c) Hanson, R. M. Chem. ReV.
1991, 91, 437. (d) Pfenninger, A. Synthesis 1986, 89. (e) Behrens, C. H.;
Sharpless, K. B. Aldrichimica Acta 1983, 16, 67.
(3) Tokiwa, Y.; Miyoshi-Saitoh, M.; Kobayashi, H.; Sunaga, R.; Konishi, M.;
Oki, T.; Iwasaki, S. J. Am. Chem. Soc. 1992, 114, 4107.
(4) Edo, K.; Mizugaki, M.; Koide, Y.; Seto, H.; Furihata, K.; Otake, N.; Ishida,
N. Tetrahedron Lett. 1985, 26, 331.
(5) Ho¨fle, G.; Bedorf, N.; Steinmetz, H.; Schomburg, D.; Gerth, K.; Reichen-
bach, H. Angew. Chem., Int. Ed. Engl. 1996, 35, 1567.
(6) (a) Katsuki, T.; Sharpless, K. B. J. Am. Chem. Soc. 1980, 102, 5974. (b)
For a review, see: Johnson, R. A.; Sharpless, K. B. In Catalytic Asymmetric
Synthesis, 2nd ed.; Ojima, I., Ed.; Wiley-VCH: New York, 2000; Chapter
6A.
(12) (a) Brandt, P.; Norrby, P. O.; Daly, A. M.; Gilheany, D. G. Chem.sEur.
J. 2002, 8, 4299. (b) Feichtinger, D.; Plattner, D. A. Chem.sEur. J. 2001,
7, 591.
9
13672
J. AM. CHEM. SOC. 2005, 127, 13672-13679
10.1021/ja051851f CCC: $30.25 © 2005 American Chemical Society

Reactivity/Selectivity in Jacobsen−Katsuki Epoxidation
A R T I C L E S
Table 1. Relative Kinetics in the Jacobsen-Katsuki Epoxidation
for the Eight Alkenes under Investigation along with Measured
Enantio- and Diastereoselectivities
tions of styrene and the relevant methyl-substituted styrene in
hand, the relative reactivity of the two substrates could be
determined, assuming a first-order reaction in each styrene and
identical reaction orders in other reactants (eq 1; A and B are
the concentrations of the two competing substrates, with
subscript “0” indicating initial concentration).
major
relative
reactivitya
ee
absolute
configurationb
diastereomeric ratio
(dr)
entry
alkene
product
epoxide
1
2
3
4
5
6
7
8
1
5
1.00
0.75
0.39
0.63
53%
89%
25%
54%
73%
71%
61%
<5%
(R)
(R)
(R)
(S)
(S)
(R)
(R)
2a
2b
2c
3a
3b
3c
4
6a
6b
6c
7a
7b
7c
g
10.3c
72.2d
A0
B0
ln
) k ln
(1)
∼500e
rel
( )
( )
0.67
A
B
0.100
0.059
0.079g
5.3f
Each reaction was first performed on the isolated alkene, and
the formation of epoxide was followed by chiral GC using
n-decane as internal standard. After workup, pure epoxides were
obtained by flash chromatography. Racemic epoxides were
synthesized separately (see Experimental Section), and the GC
burn ratios compared to n-decane were established. In this way
the selectivity of the epoxidation could be determined for each
alkene individually (Table 1).
a The correlation coefficient was larger than 0.995 in all cases. b At the
benzylic position the presence or absence of a methyl group in the R-position
does not change the sequence in the Cahn-Ingold-Prelog system, so (R)
indicates the same sense of chirality throughout. c The ee of the diastere-
omeric epoxide 6b was 73%, and the configuration was (S). d The high dr
resulted in very small amounts of the diastereomeric epoxide, and the ee
could not be determined accurately. e Only the major enantiomer of the
diastereomeric epoxide was detected. f The ee of the diastereomeric epoxide
7a was 31%, and the configuration was (S). g Mostly to non-epoxide
products.
The GC conditions used (see Experimental Section for details)
also allowed an estimation of the mass balance during the
competition experiments. The initial competition experiments
were performed under conditions identical to those of the
preparative reactions, except for the use of equimolar quantities
of styrene and the alkene under investigation. Of the eight
alkenes (Chart 1), styrene was the fastest reacting substrate and
its relative reactivity was set to unity. This first set of compe-
tition experiments gave satisfactory results for all three mono-
methyl-substituted alkenes 2a-c and one of the dimethyl-
substituted alkenes (3a). The data from these kinetic experiments
were plotted according to eq 1, yielding a straight line for each
competition experiment, as expected (Figure 1).
Chart 1. Structures of the Eight Styrenes Investigated
As shown in Table 1 (entries 2-4) the relative reactivities
of 2a-c were 0.75, 0.39, and 0.63, respectively, and in all three
cases the correlation was excellent, with r2 > 0.999. Interest-
ingly, the dimethyl-substituted 3a was the third-fastest reacting
alkene (relative reactivity 0.67, r2 ) 0.997) of all the substrates
tested (Table 1, entry 5).
The two remaining dimethyl-substituted styrenes 3b and 3c
were subjected to identical competition experiments with
styrene, and the initial results obtained within 30 min indicated
relative reactivities of 0.10 and 0.06, respectively. However,
the plot was not linear during the entire run, and both com-
pounds were therefore subjected to new competition experiments
involving the slower-reacting 2b. The measured slopes were
0.26 and 0.15, and correcting for the reactivity of 2b itself, one
again obtains relative reactivities of 0.10 and 0.06 for 3b and
3c, respectively, but now with increased accuracy (Table 1,
entries 6 and 7; r2 ) 0.997 and 0.996, respectively).
studies.12,13 In this paper, we present the results of a reactivity/
selectivity investigation of the Jacobsen-Katsuki epoxidation,
based on kinetic studies of a series of methyl-substituted
styrenes, as part of an effort to understand the factors responsible
for the orientation of the substrate in the proposed transition
state for the reaction.
Our starting point was the large difference in enantioselec-
tivity observed (see ref 7a and Table 1, entries 2 and 3) for the
epoxidation of (Z)- and (E)-â-methylstyrene, respectively. We
therefore chose to investigate the set of eight alkenes shown in
Chart 1, to systematically probe the effects of methyl substitution
by determination of (i) reactivity relative to that of styrene itself
in competition experiments, (ii) enantioselectivity and absolute
configuration, and (iii) diastereoselectivity in cases where
epimerization could be observed (2a,b, 3a,b).
Tetrasubstituted alkene 4 was very unreactive in the standard
preparative asymmetric epoxidation procedure (only 24%
conversion after 3 h), and only minor amounts of the epoxide
could be isolated, due to degradation of the reactant and product
under the reaction conditions. In the competition experiment, 4
was epoxidized together with 2b and the reactivity of 4 relative
to styrene was calculated to 0.09 with r2 ) 0.995 (Table 1,
entry 8), but again only minor amounts of the epoxide could
be detected. The plot for epoxidation of alkenes 3b, 3c, and 4
in competition with 2b is shown in Figure 2.
Results and Discussion
All the alkenes 1-4 could be separated using GC (for
retention times, see Supporting Information, Table S1), allowing
determination of the relative concentrations at various levels
of conversion in competition experiments. With the concentra-
(13) (a) Cavallo, L.; Jacobsen, H. Inorg. Chem. 2004, 43, 2175. (b) Abashkin,
Y. G.; Burt, S. K. Org. Lett. 2004, 6, 59. (c) Khavrutskii, I. V.; Musaev,
D. G.; Morokuma, K. J. Am. Chem. Soc. 2003, 125, 13879. (d) Cavallo,
L.; Jacobsen, H. J. Org. Chem. 2003, 68, 6202. (e) Jacobsen, H.; Cavallo,
L. Chem.sEur. J. 2001, 7, 800. (f) El Bahraoui, J.; Wiest, O.; Feichtinger,
D.; Plattner, D. A. Angew. Chem., Int. Ed. 2001, 40, 2073. (g) Cavallo, L.;
Jacobsen, H. Angew. Chem., Int. Ed. 2000, 39, 589. (h) Abashkin, Y. G.;
Collins, J. R.; Burt, S. K. Inorg. Chem. 2001, 40, 4040. (i) Strassner, T.;
Houk, K. N. Org. Lett. 1999, 3, 419.
The relative rates, ee values for the major product from each
run (except for epoxidation of styrene 4), and diastereomeric
ratios for nonstereospecific reactions are collected in Table 1.
9
J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005 13673

A R T I C L E S
Fristrup et al.
Figure 1. Results from the epoxidations of 2a-c and 3a in competition with styrene.
Figure 2. Results from the epoxidations of 3b-c and 4 in competition with (E)-â-methylstyrene (2b).
With these data in hand, our next task was assignment of the
absolute configuration of the major products 5-7 (Chart 2).
For 5, absolute configuration was assigned by comparison with
a commercially available sample (Sigma-Aldrich), while 6a
could be assigned by comparison with literature data (see ref
7a). For 6b, 6c, and 7b we relied on chemical correlation with
epoxides derived from chiral 1,2-diols which were themselves
the products of Sharpless asymmetric dihydroxylation of alkenes
2b, 2c, and 3a, respectively (see Experimental Section for
details). Finally, the absolute configurations of 7a and 7c were
assigned by comparison with the products of Shi asymmetric
epoxidation14 of alkenes 3a and 3c, respectively.
From the results in Table 1, no simple correlation between
reactivity and selectivity is immediately obvious, and no trends
could be discerned in a plot of selectivity (measured as % ee
for the major product) as a function of relative rate in compe-
tition with styrene (Figure 3). However, we note that if the
classical side-on approach vector is postulated,15 the major
enantiomer always comes from an approach where at least one
alkene hydrogen points toward the salen ligand.16 For the
trisubstituted substrates 3a-c as well as the cis-disubstituted
(14) Tu, Y.; Wang, Z. X.; Shi, Y. J. Am. Chem. Soc. 1996, 118, 9806.
(15) Groves, J. T.; Nemo, T. E. J. Am. Chem. Soc. 1983, 105, 5786.
(16) Brandes, B. D.; Jacobsen, E. N. J. Org. Chem. 1994, 59, 4378.
9
13674 J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005

Reactivity/Selectivity in Jacobsen−Katsuki Epoxidation
A R T I C L E S
Figure 3. Selectivity (measured as %ee) plotted against the relative reactivity (measured in competition with styrene). Positive ee values denote formation
of the (R)-configuration at the benzylic position as the major product, and negative ee values, the (S)-configuration.
Chart 2. Seven Epoxides Detected as the Expected Major
for formation of the (R) and (S) configuration, respectively, at
Products from the Epoxidation Reaction of Alkenes 1-3 with the
(R,R)-Configured Jacobsen’s Catalysta
the benzylic position of each chiral epoxide (see Table 2).
From the selectivities and reactivities shown in Table 1 and
Figure 3, the isolated reactivity leading to the two enantiomeric
products can be calculated as follows (ee in %, -100 e ee e
100):
rtot ) rR + rS
rR ) rtot(100 + ee)/200
rS ) rtot - rR
In Table 2 we have listed the isolated reactivities to either
enantiomer for each substrate, relative to the total reactivity for
styrene.
On the basis of the data shown in Tables 1 and 2, we will
now attempt to reconcile our results with previous mechanistic
suggestions concerning the Jacobsen-Katsuki reaction, with
particular emphasis on the mode of approach of the alkene to
the chiral metal-oxo complex. It is now generally accepted18,19
that the Jacobsen-Katsuki epoxidation can proceed by at least
two competing pathways (concerted or radical, the latter being
invoked to explain the nonstereospecificity observed for a
number of substrates); however, irrespective of the pathway
taken, the transition state is held to be asynchronous,11b with
an initial reaction at the â-carbon of styrene derivatives. High
steric bulk at the â-position (dimethyl-substitution) would thus
be expected to retard the rate irrespective of which enantiomer
of the epoxide is being formed (Table 1, entry 7; Table 2, entries
7 and 14).
a Trimethylstyrene (4) did not yield significant amounts of epoxide under
these reaction conditions.
substrate 2a (Table 1, entries 2 and 5-7), this condition is
sufficient to predict the absolute sense of chirality, as already
shown for related systems by the groups of Jacobsen16 and of
Katsuki.17 However, the relative reactiVity of the substrates is
still puzzling. In the earlier studies, trisubstituted alkenes were
assumed to be less reactive than cis-disubstituted ones,16,17
whereas, in the current study, the rate of epoxidation of substrate
3a is comparable to that of the favored substrate 2a (Table 1,
entries 2 and 5). Furthermore, the switch in absolute sense of
chirality between 2b and 2c (Table 1, entries 3 and 4), which
both would be required to point either a methyl group or a
phenyl group toward the salen, requires a more detailed
consideration of the steric influences of the substituents on the
possible approach vectors. New trends become apparent if the
data from Table 1 are converted to rankings of relative rates
Several approach vectors have been suggested in the literature,
differing in which part of the salen ligand is interacting with
the substrate.7c,d,13,20,21 The various suggestions have been
(18) For a review, see: Limberg, C. Angew. Chem., Int. Ed. 2003, 42, 5932.
(19) McGarrigle, E. M.; Gilheany, D. G. Chem. ReV. 2005, 105, 1563.
(17) Fukuda, T.; Irie, R.; Katsuki, T. Synlett 1995, 197.
9
J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005 13675

A R T I C L E S
Fristrup et al.
Table 2. Isolated Reactivities Leading to Either (R)- or (S)-configuration at the Benzylic Position
evaluated in a recent review.19 The common denominator,
derived from earlier work using porphyrin ligands,15 is that the
alkene approaches side-on, which in combination with the
requirement of initial attack at the â-position yields four possible
orientations for styrenes (Figure 4). Of these, the large majority
preference for either of these paths, as well as on the inherent
selectivity of each path. With the enantiomer of the metal-oxo
complex employed here the paths to the right in Figure 4 are
disfavored, so that the distal path leads predominantly to the
(R)-configuration at the benzylic position of the epoxide. This
approach can be highly enantioselective (Table 1, entry 2) but
can be blocked to a large extent by an R-methyl substituent
(Table 2, entries 4-6). A â-methyl substituent has negligible
effect on the distal approach when pointing away from the salen
(cis to the phenyl substituent) but reduces the reaction rate by
approximately a factor of 2 when pointing toward the salen
(trans). For substrates with an R-methyl, further methyl substitu-
tion at the â-position seems to have very little effect (Table 2,
entries 4 and 6).
Figure 4. Illustration of the distal phenyl and proximal phenyl approach
vectors.
The proximal phenyl path has previously only been consid-
ered as a least-hindered approach for trisubstituted alkenes in
catalytic epoxidation,16,17 but the importance of stabilizing
interactions between two aryl moieties for the selectivity in
catalytic oxidations has been amply demonstrated for the
Sharpless asymmetric dihydroxylation reaction.22 Assuming that
the approach orientation of the alkene itself is similar to that
for the distal vector, the proximal vector should lead predomi-
nantly to the (S)-configuration at the benzylic position of the
of published approach vectors have used the orientation that
we term “distal”, but both of the groups of Jacobsen16 and
Katsuki17 have also suggested that trisubstituted alkenes can add
with a phenyl substituent pointing toward the salen (here termed
“proximal” approach) when such an approach avoids the more
serious penalty of pointing two substituents toward the salen.
Separating the approaches into “distal” and “proximal” paths,
we can see that the enantioselectivity is dependent on the
(20) Dalton, C. T.; Ryan, K. M.; Wall, V. M.; Bousquet, C.; Gilheany, D. G.
Top. Catal. 1998, 5, 75.
(21) Katsuki, T. AdV. Synth. Catal. 2002, 344, 131.
(22) (a) Fristrup, P.; Tanner, D.; Norrby, P. O. Chirality 2003, 15, 360. (b)
Kolb, H. C.; Andersson, P. G.; Sharpless, K. B. J. Am. Chem. Soc. 1994,
116, 1278.
9
13676 J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005

Reactivity/Selectivity in Jacobsen−Katsuki Epoxidation
A R T I C L E S
epoxide, in good agreement with the data in Table 2. We
propose that this path is essentially blocked (sterically) for
substrates such as 2a and 3b with a â-methyl cis to the phenyl
(Table 2, entries 12 and 13) but is the major or even exclusive
pathway for substrate 3a, where the opposite alkene side is
blocked by two substituents (Table 2, entry 8). If it is assumed
that 2a reacts exclusively by the distal phenyl path, and 3a
exclusively by the proximal path, the ee values in Table 1
(entries 2 and 5) can be taken as a measure of the “inherent”
enantioselectivity of each pathway (89% for distal vs 73% for
proximal). Styrene itself should be able to react via both
pathways, and the data in Table 2 (entries 1 and 11) indicate
an approximately 3-fold difference in rate, in favor of the distal
approach. A â-methyl substituent trans to the phenyl group has
no measurable effect on the proximal path (Table 2, entries 10
and 11) but provides a weak inhibition of the distal phenyl path
(Table 2, entries 1 and 3), leading to an almost complete loss
of enantioselectivity for 2b in the traditional Jacobsen system
(Table 1, entry 3).7a
favored intermediate, and in this case the stereospecificity is
high, ca. 70:1. For the two dimethyl-substituted styrenes (3a
and 3b), the difference in energy between (E)- and (Z)-
configuration is much smaller (the â-methyl is cis to either a
methyl or a phenyl), so for these substrates the stabilization of
the radical by noncovalent interactions with the salen is expected
to be more important. For substrate 3a the intermediate radical
is already stabilized in the proximal phenyl approach, and
consequently the propensity to isomerization is low, resulting
in our highest observed stereospecificity, about 500:1. For alkene
3b both possible approaches are subject to steric hindrance, but
entry through the distal path followed by isomerization in the
intermediate radical can provide stabilization by orienting the
phenyl in a position proximal to the salen, leading to our highest
observed epimerization, almost 20% (5:1).
Summary and Conclusions
Differences in selectivity and reaction rate in the Jacobsen-
Katsuki epoxidation of a range of methyl-substituted styrenes
have been rationalized as arising from a competition of approach
vectors with opposite enantiomeric preference at the benzylic
position. One vector (termed “distal”) corresponds to most
earlier proposals of approaches, whereas a less frequently in-
voked approach with the phenyl group coming in close to the
salen is termed “proximal”. The effect of styrene substitution
on the rate through each of these vectors has been elucidated.
The current results also allow identification of the factors
needed for vector switching in the epoxidation reaction. Styrenes
with (E)-R,â-dialkyl substitution will not be able to react via a
distal approach and will therefore give high selectivities, but
with an opposite sense of chiral induction at the benzylic center
relative to that expected for the “classical” (Z)-â-alkyl-styrene
substrates and with a comparable rate. For a styrene to be a
favored substrate in the Jacobsen-Katsuki epoxidation, in the
sense that it should react both fast and with high enantioselec-
tivity, one side must be blocked by two substituents, whereas
the other side can accept a phenyl but not an alkyl substituent.
All styrene substitution patterns except the two favored ones
give either lower selectivity, lower reactivity, or both.
The Mn(salen)-catalyzed epoxidation is known to be a
strongly electrophilic reaction7c,7d,11b,e,g,h and, thus, would be
expected to be accelerated by the electron-donating methyl
substituents. For the distal path, the observed effect is instead
a retardation, due either to the steric blocking that has already
been discussed or to loss of conjugation, which penalizes the
transition state more than the ground state. The latter effect is
implied from the result for 2a (Table 2, entry 2), which reacts
slightly slower than styrene itself (Table 2, entry 1), despite
the higher electron density. For the proximal path on the other
hand, it is clear that R-methyl substitution increases the reactivity
(Table 2, entries 8 and 9), probably due to a pure inductive
effect and lack of steric blocking. The inductive effect is strong
enough to overcome most of the penalty incurred by a proximal
phenyl group. As a result, the dimethyl-substituted styrene 3a
reacts through the proximal path with a rate rivaling the two
substrates that are unhindered in the distal path, 1 and 2a,
whereas the two trisubstituted alkenes that have to react with a
methyl group pointing toward the salen are much slower (3b
and 3c, Table 1, entries 6 and 7).
The epoxidation reaction is fairly stereospecific under the
conditions employed here, in no case yielding more than 20%
epimerized product. There is a clear consensus that epimerized
products are produced through bond rotation in a radical
intermediate,7c,d strongly supported by the observation that
epimerization is completely suppressed in substrates substituted
with a radical trap.11f There are also strong indications that the
radical intermediate is produced as a side path and that the main
reaction under the conditions employed here proceeds through
a concerted, asynchronous TS.11b,f However, for the part of the
reaction proceeding through a radical intermediate, the epoxide
ring formation can occur either with or without epimerization
of the benzylic stereocenter. For the four substrates where we
can measure epimerization, the â-monomethyl-substituted sty-
renes 2a-b and 3a-b, the diastereomeric ratios can be used to
judge the relative propensity of the radical intermediates for
bond rotation. For (Z)-â-methyl styrene (2a), bond rotation in
the radical leads to the (E)-configured intermediate, which is
favored energetically, and this accounts for the observed high
degree of epimerization, approximately 10%. In the case of (E)-
â-methylstyrene (2b) the radical would isomerize to a less
Experimental Section
Glassware was dried in an oven at 140 °C. Dichloromethane,
n-hexane, and n-pentane were purchased from LabScan (HPLC grade)
and were used as received. Diethyl ether was from Bie & Berntsen
and used as received. (R,R)-(-)-N,N′-Bis(3,5-di-tert-butylsalicylidene)-
1,2-cyclohexanediaminomanganese(III) chloride, R-methylstyrene, (E)-
â-methylstyrene, styrene oxide, (R)-(+)-styrene oxide, 2-bromo-3-
methyl-2-butene, and n-decane were purchased from Aldrich and used
as received. (Z)-â-Methylstyrene was purchased from TCI (Japan) and
used as received. GC was performed on a Shimadzu GC-2010 equipped
with an AOC-20i autosampler for 12 samples. Procedure A: Chrompack
CP Chirasil-Dex CB 0.25 mm × 25 m column and temperature
program, which was 90 °C for 20 min followed by a ramp (20 °C/
min) to 180 °C which was kept for 15.50 min. Styrene oxide and
R-methylstyrene oxide were analyzed using procedure B: Supelco
â-Dex 120 0.25 mm × 30 m column and temperature program was
100 °C for 20 min followed by a ramp (20 °C/min) to 180 °C which
was kept for 15.50 min.
General Procedure for Preparative Epoxidations: The olefin
under investigation (1.0 mmol) and n-decane (0.5 mmol, 71 mg) were
dissolved in dichloromethane (1 mL) and cooled to 0 °C. After mixing,
the first sample (20 µL) was withdrawn and diluted to 1.5 mL with
9
J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005 13677

A R T I C L E S
Fristrup et al.
diethyl ether. (R,R)-(-)-N,N′-Bis(3,5-di-tert-butylsalicylidene)-1,2-cy-
clohexanediaminomanganese(III) chloride (0.04 mmol, 26 mg) was
added to the reaction vessel, and a precooled (0 °C) aqueous solution
of NaOCl (∼0.55 M, 3.5 mL) was added. 20 µL samples of the organic
phase were withdrawn after 1, 30, 60, 120, and 180 min, each time
with stirring stopped briefly to allow phase separation. The sample was
filtered through 1 cm of silica using 1.5 mL of diethyl ether and
analyzed by chiral GC. After 3 h the reaction mixture was diluted with
hexane (25 mL), and the phases separated. The organic phase was
washed once with H2O (30 mL) and twice with brine (30 mL), then
dried over Na2SO4, filtered, and evaporated to give a yellow oil. The
crude product was purified by flash chromatography (5% ether in
(Z)-â-Methylstyrene oxide was isolated as a clear oil in 86% yield
(2.31 g). GC retention times were 15.8 min and 18.4 min, respectively,
for the two enantiomers (procedure A).
1H NMR (300 MHz, CDCl3, δ): 1.09 (dd, 5.4 Hz, 0.9 Hz, 3H),
3.31-3.39 (m, 1H), 4.07 (d, 4.2 Hz, 1H), 7.25-7.39 (m, 5H).26 13C
NMR (75.4 MHz, CDCl3, δ): 12.79, 55.38, 57.77, 126.81, 127.71,
128.24, 135.75.27
R-Methylstyrene oxide was isolated as a clear oil in 77% yield (1.03
g, 10 mmol scale). GC retention times were 18.4 min and 18.8 min,
respectively, for the two enantiomers (procedure B).
1H NMR (300 MHz, CDCl3, δ): 1.72 (s, 3H), 2.80 (d, 5.4 Hz, 2H),
2.97 (d, 5.4 Hz, 2H), 7.21-7.40 (m, 5H).28 13C NMR (75.4 MHz,
CDCl3, δ): 21.77, 56.73, 57.05, 125.25, 127.42, 128.30, 141.10.29
â,â-Dimethylstyrene oxide was isolated as a clear oil in 81% yield
(2.41 g). GC retention times were 14.7 min and 15.0 min, respectively,
for the two enantiomers (procedure A).
1H NMR (300 MHz, CDCl3, δ): 1.08 (s, 3H), 1.49 (s, 3H), 3.87 (s,
1H), 7.24-7.39 (m, 5H).30 13C NMR (75.4 MHz, CDCl3, δ): 18.2,
25.0, 61.3, 64.8, 126.6, 127.6, 128.3, 136.8.30
(E)-R,â-Dimethylstyrene oxide was isolated as a clear oil in 55%
yield (1.64 g). GC retention times were 17.3 min and 17.6 min,
respectively, for the two enantiomers (procedure A).
1H NMR (300 MHz, CDCl3, δ): 1.43 (d, 5.4 Hz, 3H), 1.67 (s, 3H),
2.95 (q, 5.5 Hz, 1H), 7.22-7.38 (m, 5H).31 13C NMR (75.4 MHz,
CDCl3, δ): 14.7, 17.6, 60.6, 62.8, 125.3, 127.4, 128.5, 143.3.
(Z)-R,â-Dimethylstyrene oxide was isolated as a clear oil in 79%
yield (2.29 g). GC retention times were 12.8 min and 13.8 min,
respectively, for the two enantiomers (procedure A).
1H NMR (300 MHz, CDCl3, δ): 0.98 (dd, 5.4 Hz, 0.9 Hz, 3H),
1.64 (s, 3H), 3.17 (q, 5.4 Hz, 1H), 7.23-7.38 (m, 5H).32 13C NMR
(75.4 MHz, CDCl3, δ): 14.7, 24.8, 61.5, 62.9, 126.8, 127.3, 128.3,
139.9.
1
pentane). The products were characterized by H NMR spectroscopy
(see below) and analyzed by chiral GC.
General Procedure for Competitive Epoxidations: Styrene (0.5
mmol, 51 mg), n-decane (0.5 mmol, 71 mg), and the olefin under
investigation (0.5 mmol) were dissolved in dichloromethane (1 mL)
and cooled to 0 °C. After mixing, the first sample (20 µL) was
withdrawn and diluted to 1.5 mL with diethyl ether. (R,R)-(-)-N,N′-
Bis(3,5-di-tert-butylsalicylidene)-1,2-cyclohexanediaminomanganese-
(III) chloride (0.04 mmol, 26 mg) was added to the reaction vessel,
and a sample (20 µL) was withdrawn and filtered through 1 cm of
silica using 1.5 mL of diethyl ether. A precooled (0 °C) aqueous solution
of NaOCl (∼0.55 M, 3.5 mL) was added with rapid stirring. 20 µL
samples of the organic phase were withdrawn after 2.5, 5, 7.5, 10, 15,
30, 45, 60, 90, and 120 min and filtered as mentioned above, each
time stopping the stirring for 10 s to allow phase separation. All samples
were analyzed by chiral GC using the appropriate general temperature
program.
The competitive epoxidations of compounds 3b-c and 4 were
performed using (E)-â-methylstyrene (0.5 mmol, 56 mg) instead of
styrene as the only change.
Dimethyl-substituted styrenes 3 were synthesized according to
literature procedures.22a R,â,â-Trimethylstyrene was synthesized via a
similar Suzuki coupling procedure with 2-bromo-3-methyl-2-butene as
the bromide coupling partner. The reaction gave the desired product
in 87% yield (4.74 g) after chromatography (pentane).
R,â,â-Trimethylstyrene oxide (5 mmol scale) was isolated as a clear
oil in 87% yield (0.71 g). GC retention times were 14.2 min and 15.6
min, respectively, for the two enantiomers (procedure A).
1H NMR (300 MHz, CDCl3, δ): 0.97 (s, 3H), 1.48 (s, 3H), 1.62 (s,
3H), 7.20-7.38 (m, 5H).33 13C NMR (75.4 MHz, CDCl3, δ): 21.0,
21.6, 21.9, 64.0, 66.8, 126.3, 127.0, 128.2, 142.5.
1H NMR (300 MHz, CDCl3, δ):1.58 (br s, 3H), 1.80 (br s, 3H),
1.95 (br s, 3H), 7.06-7.33 (m, 5H).23 13C NMR (CDCl3, 75.4 MHz,
δ): 20.8, 21.1, 22.3, 54.7, 126.0, 127.5, 128.2, 128.7, 130.3, 145.6.
Racemic (E)-â-Methylstyrene oxide was synthesized using a one-
pot procedure developed by Kolb and Sharpless.24 The desired epoxide
6b was isolated as a clear oil in 60% overall yield (2.15 g). GC retention
times were 13.8 min and 14.0 min, respectively, for the two enantiomers
(procedure A).
Synthesis of enantioenriched epoxides a by three-step procedure
(dihydroxylation, tosylation, and ring-closure):
R-Methylstyrene (10 mmol, 1.18 g) was added to 100 mL of a
solution of K2OsO2(OH)4 (0.05 mmol, 18 mg), (DHQD)2PHAL (0.10
mmol, 78 mg), K3Fe(CN)6 (30 mmol, 9.88 g), and K2CO3 (30 mmol,
4.14 g) in t-BuOH/H2O, 1:1 cooled to 0 °C. The resulting mixture was
stirred at 0 °C overnight and then quenched by addition of Na2S2O3
(10 g). The phases were separated and the water phase was extracted
three times with EtOAc (100 mL). The combined organic phases were
washed with 5% H2SO4 (20 mL), saturated aqueous NaHCO3 (60 mL),
and then saturated aqueous NaCl (60 mL), dried over MgSO4, and
purified by flash chromatography on silica gel (40% EtOAc in heptane)
to give the diol in quantitative yield (1.52 g).
1H NMR (300 MHz, CDCl3, δ): 1.45 (br d, 5.4 Hz, 3H), 3.04 (qd,
2.1 Hz, 5.4 Hz, 1H), 3.58 (br d, 2.1 Hz, 1H), 7.23-7.38 (m, 5H).25 13
C
NMR (75.4 MHz, CDCl3, δ): 16.9, 58.0, 58.5, 124.5, 127.0, 127.4,
136.7.25
The racemic epoxides 6a, 6c, 7a-c, and 8 were all synthesized
according to this general procedure: The alkene (1.0 equiv, 20 mmol)
was dissolved in CH2Cl2 (100 mL), and the solution was cooled to 0
°C (ice-bath). MCPBA (1.1 equiv, 22 mmol, 4.93 g) was added slowly
with stirring. The reaction mixture was left stirring for 2 h and then
transferred to a separatory funnel with CH2Cl2 (20 mL). n-Pentane (500
mL) was added, and the organic phase was extracted twice with
saturated aqueous NaHCO3 (100 mL) and then saturated aqueous NaCl
(100 mL) and dried over MgSO4. The solvent was removed on a rotary
evaporator (12 mmHg, 25 °C), and the crude product was purified by
flash column chromatography (5% ether in pentane). A solution of
phosphomolybdic acid (12 g) in ethanol (250 mL) was used to visualize
the product on TLC.
The diol (1 mmol, 152 mg) was dissolved in CH2Cl2 (2 mL), and 1
equiv of TsCl (192 mg) was added followed by the addition of 2 equiv
of Et3N. The solution was stirred at rt overnight. Pentane (∼50 mL)
(26) Sasaki, H.; Irie, R.; Hamada, T.; Suzuki, K.; Katsuki, T. Tetrahedron 1994,
50, 11827.
(27) Monti, J. P.; Faure, R.; Sauleau, A.; Sauleau, J. Magn. Reson. Chem. 1986,
24, 15.
(28) Archelas, A.; Furstoss, R. J. Org. Chem. 1999, 64, 6112.
(29) Concello´n, J. M.; Cuervo, H.; Ferna´ndez-Fano, R. Tetrahedron 2001, 57,
8983.
(30) Pedragosa-Moreau, S.; Archelas, A.; Furstoss, R. Tetrahedron 1996, 52,
4593.
(31) Wang, Z.-X.; Tu, Y.; Frohn, M.; Zhang, J.-R.; Shi, Y. J. Am. Chem. Soc.
1997, 46, 11224.
(23) Berthiol, F.; Doucet, H.; Santelli, M. Eur. J. Org. Chem. 2003, 6, 1091.
(24) Kolb, H. C.; Sharpless, K. B. Tetrahedron 1992, 48, 10515.
(25) Besse, P.; Renard, M. F.; Veschambre, H. Tetrahedron: Asymmetry 1994,
7, 1249.
(32) Satoh, T.; Kobayashi, S.; Nakanishi, S.; Horiguchi, K.; Irisa, S. Tetrahedron
1999, 55, 2515.
(33) Yamamoto, H.; Miura, M.; Nojima, M.; Kusabayashi, S. J. Chem. Soc.,
Perkin Trans. 1 1986, 173.
9
13678 J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005

Reactivity/Selectivity in Jacobsen−Katsuki Epoxidation
A R T I C L E S
was added, and the mixture was washed with saturated aqueous NaCl
(5 mL). The crude epoxide was purified by flash chromatography on
silica gel (1% Et3N, 10% ether in pentane) to give 3c in 23% yield.
The enantiomeric excess was determined to be 95% by chiral GC using
procedure B.
(E)-R,â-Dimethylstyrene (5 mmol, 661 mg) was added to a solution
of K2OsO2(OH)4 (0.025 mmol, 9 mg), (DHQD)2PHAL (0.05 mmol,
39 mg), K3Fe(CN)6 (15 mmol, 4.94 g), and K2CO3 (15 mmol, 2.07 g)
in t-BuOH/H2O, 1:1 (50 mL) cooled to 0 °C. The resulting mixture
was stirred at 0 °C overnight and then quenched by addition of 5 g of
Na2S2O3. The phases were separated, and the water phase was extracted
three times with EtOAc (50 mL). The combined organic phases were
washed with 5% H2SO4 (10 mL), saturated aqueous NaHCO3 (30 mL),
and then saturated aqueous NaCl (30 mL), dried over MgSO4, and
purified by flash chromatography on silica gel (40% EtOAc in heptane)
to give the diol in 66% yield (547 mg).
amount of DMAP was added, and after refluxing for 96 h the reaction
was worked up. The reaction mixture was transferred to a separatory
funnel with CH2Cl2 (10 mL), diluted with pentane (100 mL), and
washed with saturated aqueous NaCl (10 mL). The crude epoxide was
purified by flash chromatography on silica gel (1% Et3N, 10% ether in
pentane) to give 7b in 54% yield. The enantiomeric excess was
determined to be 95% by chiral GC using procedure A.
Acknowledgment. We wish to thank Professor Declan
Gilheany for illuminating discussions. Professor Saverio Florio
kindly provided suggestions for the chiral analysis of R-meth-
ylstyrene oxide. Financial support from the Danish Technical
Sciences Research Council and the Technical University of
Denmark is gratefully acknowledged.
Supporting Information Available: GC retention times for
alkenes, epoxides, and byproducts. This material is available
free of charge via the Internet at http://pubs.acs.org.
The diol (0.5 mmol, 83 mg) was dissolved in CH2Cl2 (5 mL), and
1 equiv of TsCl (95 mg) was added followed by the addition of Et3N
(1 mL). The solution was refluxed overnight and became yellow, but
TLC showed only limited conversion of the diol. A small (∼5 mg)
JA051851F
9
J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005 13679