Enthalpy- and/or entropy-controlled asymmetric oxidation: Stereocontrolling factors in Mn-salen-catalyzed oxidation

Article abstract of DOI:10.1016/S0040-4039(00)01210-7

The degree of enantioselection by second-generation Mn-salen complexes was found to depend upon the conformation of their ligand and substrate nucleophilicity. Oxidation of usual olefins was better effected by using (R,S)-Mn-salen complexes as catalysts, while that of more nucleophilic ones was achieved by using (R,R)-Mn-salen complexes. This phenomenon was explained by analyzing the enthalpy and entropy factors of the reactions. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)01210-7

Tetrahedron Letters 41 (2000) 7053±7058
Enthalpy- and/or entropy-controlled asymmetric oxidation:
stereocontrolling factors in Mn±salen-catalyzed oxidation
Tomohiro Nishida, Akio Miyafuji, Yoshio N. Ito and Tsutomu Katsuki*
Department of Chemistry, Graduate School of Science, Kyushu University 33, Hakozaki, Higashi-ku,
Fukuoka 812-8581, Japan
Received 1 May 2000; revised 13 July 2000; accepted 14 July 2000
Abstract
The degree of enantioselection by second-generation Mn±salen complexes was found to depend upon the
conformation of their ligand and substrate nucleophilicity. Oxidation of usual ole®ns was better e€ected by
using (R,S)-Mn±salen complexes as catalysts, while that of more nucleophilic ones was achieved by using
(R,R)-Mn±salen complexes. This phenomenon was explained by analyzing the enthalpy and entropy
factors of the reactions. # 2000 Elsevier Science Ltd. All rights reserved.
Keywords: (salen)manganese(III) complex; asymmetric oxidation; entropy e€ect.
Optically active (salen)manganese(III) complexes (hereafter referred to as Mn±salen complexes)
1
are potent catalysts for enantioselective oxidation. Mn±salen complexes which carry chiral units
at both the ethylenediamine and salicylaldehyde parts, especially second-generation Mn±salen
complexes bearing a binaphthyl unit as the chiral auxiliary at the salicylaldehyde part, exhibit
2
versatile and ecient asymmetric catalysis: epoxidation, oxidation of enol ethers, sulfoxidation,
3
sul®midation, aziridination, kinetic resolution (KR) of racemic allenes, benzylic oxidation,
desymmetrization of meso-heterocycles, etc. were e€ectively performed in a highly enantio-
selective manner. Since second-generation Mn±salen complexes have two di€erent chiral units,
they are classi®ed into two diastereomeric complexes, (R,S)- and (R,R)-complexes. Among the
4
above reactions, epoxidation, oxidation of enol ethers, and KR of racemic allenes were eciently
e€ected by (R,S)-complexes in the presence of a donor ligand like 4-phenylpyridine N-oxide,
while sulfoxidation and desymmetrization of meso-heterocycles are e€ected by (R,R)-complexes
in the absence of a donor ligand. The role of the donor ligand was clari®ed by X-ray di€raction
5
analysis of the diastereomeric Mn±salen complexes (1 and 2). However, the question why the
best catalyst for each oxidation varies with the substrate has remained unanswered.
*
Corresponding author. Fax: +81-92-642-2607; e-mail: katsuscc@mbox.nc.kyushu-u.ac.jp
0040-4039/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)01210-7

7
054
Although (R,S)- and (R,R)-complexes (1 and 2) bearing two aqua ligands at their apical
5
positions are diastereomeric, their ligand conformations are considerably di€erent (Fig. 1). The
basal salen ligand of 1 takes a slightly folded stepped conformation, while the ligand of 2 takes a
much folded stepped conformation. The oxo Mn(V) species, which is the active intermediate in
Mn(III)±salen-catalyzed oxidation, is also considered to possess a ligand conformation similar to
5
that of the parent Mn(III)-complexes, respectively. Substrates are expected to approach the oxo
Mn(V) species beyond the downward naphthalene ring of the basal salen ligand. As indicated by
arrows in Fig. 1, therefore, substrates approach the oxo Mn(V) species from the left side along
path a in the reaction with complex 1 as the catalyst, and from the right side along path b in the
reaction with complex 2 as the catalyst, if the oxene atom of the oxo species is delivered on the
6
top side of the complexes. As also shown in Fig. 1, the top apical aqua ligand, which is replaced
with an oxene atom upon oxidation to the corresponding oxo species, exists in a concavity
0
0
surrounded by the four gray-colored substituents on the ligand. However, in path a, the 2 -phenyl
group, which is located closest to the oxene atom and causes the strongest steric repulsion and
p±p anti-bonding interactions with the incoming ole®ns, protrudes in front of the oxene atom,
while, in path b, it exists in the innermost position of the path. The critical distance for formation of
a charge transfer complex between ole®ns and an oxo metal species has been reported to be ca. 3.5
Ê
7
Ê
A or less. The tip of the arrow in Fig. 1 is placed at a distance of 3.5 A from the corresponding
0
0
aqua ligand. Thus, the incoming ole®n along path a interacts with the 2 -phenyl group at an early
stage of the reaction and the repulsion between them plays an important role in the determination
Figure 1.

7
055
of the orientation of the incoming ole®n and, in turn, the sense of enantioselection by the complex.
Accordingly, the oxidation with complex 1 is expected to be enthalpy dependent. On the other
0
0
hand, the ole®n along path b should interact with the 2 -phenyl group at a late stage of the reaction
and the enantioselectivity of the reaction should be in¯uenced not only by the weak interaction
between the salen ligand and the incoming ole®n but also by the di€erence in the frequency factor
in respect of the transit of the ole®ns of di€erent orientations through the area surrounded by the
four groups, disordering the solvent arrangement around the concavity. This latter e€ect is related
to the entropy factor, and the reaction with 2 is considered to be more entropy dependent than that
with 1. From these analyses, it was expected that epoxidation of usual ole®ns would be better
e€ected by using complex 1 as the catalyst, but that complex 2 might be a superior catalyst for the
reaction of a more nucleophilic ole®n in which the distance between the ole®n and oxene atom at
0
0
the transition state should be elongated, and the repulsion between the substrate and the 2 -phenyl
group becomes smaller and the contribution of the enthalpy factor to enantioselectivity is reduced.
To explore this possibility, we examined oxidation of cis-ole®ns, indene 5 and more nucleophilic
ole®ns like N-toluenesulfonyl-1,2,3,4-tetrahydropyridine (NTTP) 6 and 3,4-dihydro-2H-pyrane
8
(
DHP) 7, with complexes 1±4 (Table 1). Reaction was carried out in ethanol, with iodosylbenzene
9
as the oxidant in the absence of a donor ligand. The reaction of 6 gave the corresponding
mixture of cis- and trans-b-hydroxy-a-ethoxypyrrolidines, and that of 7 gave trans-2-hydroxy-1-
ethoxytetrahydrofurans. Oxidation of indene gave the corresponding epoxide when the reaction
ꢀ
was carried out at a temperature lower than 20 C, while a mixture of 2-hydroxy-1-ethoxyindane
ꢀ
and indeneoxide was obtained by the reaction at the temperature above 40 C. The enantiomeric
excesses of the respective cis- and trans-products were found to be identical. As we expected,
(R,S)-complex 1 or 3 was a superior catalyst to (R,R)-complex 2 or 4, respectively, when the
substrate was 5 (entries 1 and 2), while (R,R)-complex 2 or 4 was a superior catalyst for oxidation
of more nucleophilic substrates, 6 and 7 (entries 3±6).10
Table 1
Asymmetric oxidation with complexes 1±4 as the catalysts
a)

7
056
Enantioselectivity of the asymmetric reaction is given by the Eyring equation: ln(k /
R
{
{
k )=^ÁÁH /RT+ÁÁS . As discussed above, the enantioselectivity observed with complex 3 as
S
the catalyst should be enthalpy dependent, and that with complex 4 should be a€ected by both
enthalpy and entropy factors. Therefore, the value of the enthalpy fragment of the Eyring
equation obtained with complex 3 is expected to be larger than that obtained with complex 4. It is
also expected that oxidation of more nucleophilic ole®ns would show smaller dependence on the
enthalpy fragment. Based on these analyses, we examined the reactions of 5 and 7 at various
temperatures. The results are shown in Fig. 2: ln(k /k ) was plotted against 1/T. The Eyring plots
R
S
for the oxidation of 5 with both complexes 3 and 4 sloped more steeply than the plots for the
oxidation of 7 and, of the plots for the reaction of indene, the plot obtained with complex 3 was
steeper than that obtained with complex 4. On the other hand, the plot for the reaction of 7 with
complex 4 sloped slightly less than the plot obtained with complex 3, but the enantioselectivity of
the former reaction was slightly better than that of the latter reaction. These results are all
compatible with the above discussions and suggest that the entropy factor plays an important role
in enantioselection by (R,R)-complex 4, especially when the substrate nucleophilicity is high. In
accord with these results, the enantioselectivity observed in the oxidation of 5 with (R,S)-complex
1
the enantioselectivity of the oxidation of 6 with (R,R)-complex 2 as the catalyst was much less
also showed the same temperature dependence as that with (R,S)-complex 3. On the other hand,
temperature dependent. It is noteworthy that some Eyring plots showed non-linear relationship.11
Figure 2.
In connection with this study, we have reported that complex 4 is a better catalyst for desym-
metrization of meso-tetrahydrofurans via enantiotopic selective C±H oxidation than complex 3.12
The C±H bonding orbital is, however, less nucleophilic than the p-bonding orbital. Thus, we

7
057
examined the kinetic isotopic e€ect (k /k ) in the desymmetrization to clarify the reaction
H
D
mechanism of the desymmetrization. The kinetic isotopic e€ect in hydrogen atom abstraction by
oxo metal species falls in the range of ca. 5±12.13 However, the kinetic isotopic e€ect in the
oxidation of 8-oxa-bicyclo[4.3.0]nonane and its 7,7,9,9-d derivative was 2.3. This low value
4
indicates that the desymmetrization does not proceed through hydrogen atom abstraction, but
probably through an electron transfer process. It has been reported that an electron transfer
reaction can occur at longer distance than formation of a charge transfer complex.7 These results
may explain why (R,R)-complexes are suitable catalysts for desymmetrization of meso-hetrocycles.12
Typical experimental procedure: to a solution of complex (1, 2, 3 or 4; 2.0 mg, 2.0 mmol) and
ole®n (5, 6 or 7; 0.1 mmol) in ethanol was added iodosylbenzene (22.0 mg, 0.1 mmol) at appropriate
temperature under nitrogen atmosphere and the whole mixture was stirred for 1 h (for 5), 6 h (for
b
6) and 1 h (for 7), respectively, and ®ltered through a pad of Celite and silica gel. The ®ltrate was
concentrated and the residue was chromatographed on silica gel to give the corresponding epoxide
or hydroxy acetal. The enantiomeric excesses of the products were determined by HPLC analysis
(see the footnote to Table 1). Each reaction was repeated ®ve times. All the ee's obtained at the
same temperature were within þ1% ee (þ2% ee only for the reaction of 7 with 4).
In conclusion, we were able to demonstrate that the enthalpy factor plays an important role in
stereocontrol of the oxidation with (R,S)-Mn±salen complexes, while the role of the entropy
factor becomes more pronounced in the oxidation with (R,R)-Mn±salen complexes. It is
noteworthy that second-generation Mn±salen catalysts are endowed with wide applicability to a
variety of reactions due to the fact that the reactions with diastereomeric (R,S)- and (R,R)-Mn±salen
complexes depend on enthalpy and entropy factors to di€erent degrees.
Acknowledgements
Financial support from a Grant-in-Aid for Scienti®c Research from the Ministry of Education,
Science, Sports, and Culture, Japan, is gratefully acknowledged.
References
1
. (a) Katsuki, T. Coord. Chem. Rev. 1995, 140, 189±214. (b) Katsuki, T. J. Synth. Org. Chem. Jpn. 1995, 53, 940±
51. (c) Jacobsen, E. N. In Catalytic Asymmetric Synthesis; Ojima, I., Ed.; VCH: New York, 1993; pp. 159±202.
9
. Ito, Y. N.; Katsuki, T. Bull. Chem. Soc. Jpn. 1999, 72, 603±619.
2
3
4
5
. Nishikori, H.; Ohta, C.; Oberlin, E.; Irie, R.; Katsuki, T. Tetrahedron 1999, 55, 13937±13946.
. Fukuda, T.; Katsuki, T. Tetrahedron Lett. 1996, 37, 4389±4392.
. (a) Hashihayata, T.; Punniyamurthy, T.; Irie, R.; Katsuki, T.; Akita, M.; Moro-oka, Y. Tetrahedron 1999, 55,
1
4599±14610. (b) Punniyamurthy, T.; Irie, R.; Katsuki, T.; Akita, M.; Moro-oka, Y. Synlett 1999, 1049±1052.
. For ole®n's approach to oxo Mn±salen species, see also: Kuroki, T.; Hamada, T.; Katsuki, T. Chem. Lett. 1995,
39±340.
. (a) Shimada, K.; Szwarc, M. J. Am. Chem. Soc. 1975, 97, 3321±3323. (b) He, G.-X.; Mei, H.-Y.; Bruice, T. C. Ibid.
991, 113, 5644±5650.
. Second-generation Mn±salen complexes bearing a cyclohexanediamine unit also have a basal ligand conformation
similar to the complexes bearing a diphenylethylenediamine unit: (a) Ref. 3; (b) Ohta, C.; Irie, R.; Katsuki, T.
Unpublished result.
6
7
8
3
1
9
. Oxidation of enol ethers under usual conditions gives a-hydroxy carbonyl compounds which racemize slowly
under the conditions. However, the reactions in ethanol provide the corresponding a-hydroxy acetal and aminal,
respectively, which tolerate the reaction conditions.4

7
058
1
0. (R,S)-Complex 2 was a catalyst of choice for oxidation of 1-alkoxycycloalkenes which are as nucleophilic as 7.4
However, they approach an oxo species, directing the alkoxy group downward, which su€ers steric and n±p anti-
bonding interactions. Thus, the reaction is expected to be enthalpy dependent and eciently e€ected by complex 2.
1. (a) Hamada, T.; Fukuda, T.; Imanishi, H.; Katsuki, T. Tetrahedron 1996, 52, 515±530. (b) Norrby, P.-O.; Linde,
1
Ê
C.; Akermark, B. J. Am. Chem. Soc. 1995, 117, 11035. We and others have proposed the participation of
metallaoxetane intermediate in Mn±salen-catalyzed epoxidation based on the non-linear relationship and
calculation, respectively. However, a recent study did not support its participation: Delmonte, A. J.; Haller, J.;
Houk, K. N.; Sharpless, K. B.; Singleton, D. A.; Strassner, T.; Thomas, A. A. J. Am. Chem. Soc. 1997, 119, 9907±
9
2. (a) Miyafuji, A.; Katsuki, T. Synlett 1997, 836±838. (b) Idem, Tetrahedron 1998, 54, 10339±10348.
908. At this moment, it is unclear what factor causes the non-linear relationship at lower reaction temperatures.
1
1
3. (a) Groves, J. T.; Viski, P. J. Am. Chem. Soc. 1989, 111, 8537±8538. (b) Idem, J. Org. Chem. 1990, 55, 3628±3634.
(
c) Hamada, T.; Irie, R.; Mihara, J.; Hamachi, K.; Katsuki, T. Tetrahedron 1998, 54, 10017±10028.