Origin of the asymmetric induction in metallosalen-catalyzed reactions of aldehydes

Article abstract of DOI:10.1039/b910473a

This study investigates, experimentally and computationally, the role played by the structure of the diamine moiety of salen chromium complexes in determining the catalyst conformational equilibrium and, in consequence, the direction of asymmetric induction in metallosalen-catalyzed reactions of aldehydes. The Royal Society of Chemistry 2009.

Full text of DOI:10.1039/b910473a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Origin of the asymmetric induction in metallosalen-catalyzed
reactions of aldehydesw
Wojciech Cha$adaja and Janusz Jurczak*ab
Received (in College Park, MD, USA) 28th May 2009, Accepted 17th September 2009
First published as an Advance Article on the web 7th October 2009
DOI: 10.1039/b910473a
This study investigates, experimentally and computationally, the
role played by the structure of the diamine moiety of salen
chromium complexes in determining the catalyst conformational
equilibrium and, in consequence, the direction of asymmetric
induction in metallosalen-catalyzed reactions of aldehydes.
authors6b,13 as processes catalyzed by classic Jacobsen
complex 1a.
Catalyst 1a, based on rigid (R,R)-1,2-diamine-cyclohexane,
afforded products of addition preferentially from the re face of
the aldehyde, which was obviously in agreement with the
proposed model. Similar results, albeit with lower enantio-
selectivities, were obtained for 1b, derived from (R,R)-1,2-
diphenylethylenediamine. Catalyst 1c, bearing small methyl
substituents, afforded products with even lower selectivities.
This appears logical, since the steric interaction of a small
substituent with an apical ligand would be diminished. The
direction of asymmetric induction, however, was reversed in
two out of three cases. Catalysts 1d and 1e, with significantly
larger substituents, produced better results, but also with
the opposite sense of asymmetric induction—namely (R,R)
catalyst favored addition from the si face of the substrate.
This thus raises the question of whether the whole model is
incorrect or whether the assumption of the position of the
catalyst’s conformational equilibrium fails. To establish the
position of the equilibrium and therefore answer this question,
we undertook DFT calculations.
Metallosalen complexes have proved their catalytic efficiency
in a number of asymmetric processes, including epoxidation of
olefins,1 desymmetrization or kinetic resolution of epoxides,2
as well as nucleophilic additions and cycloadditions to
aldehydes.3 In contrast to the mechanism of epoxidation
reaction,4 the mechanistic aspects of the activation of carbonyl
compounds by metallosalen catalysts, in particular the origin
of asymmetric induction, have nevertheless remained largely
investigated. Recently, a stereochemical model of metallosalen-
catalyzed transformations of aldehydes has been proposed by
Rawal5 and by the present authors6 (Scheme 1), by analogy to
Katsuki’s model of the epoxidation of olefins.7,8 Rawal also
proposed an earlier, dubious model assuming that the key role
is played by the interaction of the approaching reagent with
axial hydrogens of the cyclohexanediamine moiety.9
Three models of (Salen)Cr(III) with coordinated substrate
were compared at the B3LYP/6-31G(d) level of theory
(Table 2, entries 1–3). The most sophisticated (model I) was
the almost complete (without substituents at 5,50) catalyst 1c,
with two molecules of formaldehyde coordinated. It seemed to
reproduce the conformational equilibrium comparably to the
simpler models II and III: one lacking 3 and 30 bulky groups
and even acacen complex (II and III, respectively).
The current model assumes that the chiral diamine shifts the
conformational equilibrium of the aldehyde–catalyst complex
towards one of two possible stepped conformations in which
one side of the coordinated aldehyde is better shielded by the
benzene ring of the salicylidene moiety. According to Katsuki,7,8
substituents on the ethylenediamine moiety of the catalyst
adopt equatorial positions to avoid steric interactions with the
apical ligands of catalysts.10 This working model is supported
only by the solid-state geometry of the aldehyde–catalyst
complex, but to the best of our knowledge almost no mechanistic
studies, experimental or theoretical, have been undertaken.
The sole theoretical paper concerning salen-catalyzed cyclo-
addition to aldehydes focuses on analyzing the reaction path,
placing the emphasis on the multiplicity of the catalytic system
rather than on the stereochemical outcome.11
Moreover the structure of coordinated aldehyde seems to
have little if any influence on the discussed equilibrium.
Substitution of formaldehyde in model I with the considerably
larger acetaldehyde, benzaldehyde, or even pivaldehyde
resulted in virtually similar axial conformation preference.
Other functionals as well as the ab initio MP2 method gave
results comparable to the most popular B3LYP (entries 4–8).
To verify this model (Scheme 1) and evaluate the influence
of the structure of the catalyst’s diamine moiety on the
stereochemical course of the reaction, a series of salenCr(III)
catalysts (Fig. 1) was tested in three model reactions
(Table 1), previously reported by Jacobsen12 or by the present
a Institute of Organic Chemistry, Polish Academy of Sciences,
01-224 Warsaw, Poland. E-mail: jurczak@icho.edu.pl;
Fax: +48 22 632 66 81; Tel: +48 22 343 23 30
b Department of Chemistry, Warsaw University, 02-093 Warsaw,
Poland
w Electronic supplementary information (ESI) available: Experimental
procedures, analytical data, and computational details. See DOI:
10.1039/b910473a
Scheme 1 Stereochemical model of metallosalen-catalyzed reactions
of aldehydes.
ꢀc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 6747–6749 | 6747

View Article Online
Although the calculated absolute energy of the investigated
complexes depended strongly on the basis set, the relative
energy of their conformers seemed to be only slightly
influenced by the basis set size; even the very small 3-21G set
gave reasonable results (cf. entry 3 with 9 and 10). The
calculation employing larger basis sets, and thus the more
computationally expensive one, produced results similar to
6-31G(d) (entries 11–13).
Fig. 1 Salen catalysts variously substituted on the diamine moiety.
Table 1 Stereochemical outcome of model reactions
In view of the above observations, conformation equilibria
of catalysts 1b–1e with coordinated substrates were investi-
gated as model III at the B3LYP/6-31G(d) level of theory.
The data collected in Table 3 and depicted in Fig. 2 clearly
show that the phenyl substituent at the catalyst’s diamine
moiety preferred equatorial orientation, contrary to the alkyl
substituents. In a series of alkyl substituents the axial
conformation preference increased with the size of the group,
as a result of their mutual sterical interaction in equatorial
orientation. The preference for axial conformation of salen
chromium complex derived from 1,2-di-tert-butylethylene-
diamine (differing from 1e only in terms of a counterion)
state has been previously demonstrated in solid state.14
Contrastingly, flat phenyl rings lack such interactions and
may easily sit at equatorial positions.
Having in hand the position of the conformational
equilibrium of catalyst with coordinated aldehyde, the
Entry
Cat.
Rxn 1a
Rxn 2a
Rxn 3a
1
2
3
4
5
1a
1b
1c
1d
1e
62% (62%)
15% (53%)
9% (64%)
61% (82%)
39% (83%)
ꢁ17% (89%)
ꢁ59% (90%)
ꢁ41% (77%)
85% (84%)
34% (72%)
ꢁ9% (86%)
ꢁ34% (23%)
ꢁ7% (4%)
ꢁ47% (71%)
ꢁ90% (79%)
a
ee of major product. If the (R,R)-catalyst is used, positive and
negative values correspond to (R)- and (S)-products, respectively.
Yields are given in brackets.
Table 2 Influence of model complexity and the applied theory level
on the conformational equilibrium of the complex of 1c and aldehyde
Entry
Model
Level of theory
DEa
ꢁ8.5
1
2
3
4
5
6
7
8
I
II
B3LYP/6-31G(d)
B3LYP/6-31G(d)
B3LYP/6-31G(d)
BLYP/6-31G(d)
B3PW91/6-31G(d)
MPW1K/6-31G(d)
M05/6-31G(d)
MP2(full)/6-31G(d)
B3LYP/3-21G
B3LYP/3-21G//6-31G(d)
B3LYP/6-31G(d)+
B3LYP/6-311G(d)
B3LYP/6-311G(d)+
ꢁ8.9
ꢁ8.7
ꢁ8.5
ꢁ6.7
ꢁ7.4
ꢁ9.7
ꢁ11.0
ꢁ10.9
ꢁ7.7
ꢁ8.1
ꢁ9.1
ꢁ8.1
III
III
III
III
III
III
III
III
III
III
III
9
10
11
12
13
a
Energy difference between axial and equatorial conformation,
in kJ molꢁ1
Fig. 2 Geometries of complexes of formaldehyde with catalysts 1b–1e
.
in both axial and equatorial conformation calculated at B3LYP/6-31G(d).
ꢀc
This journal is The Royal Society of Chemistry 2009
6748 | Chem. Commun., 2009, 6747–6749

View Article Online
Table 3 Energy differences of the axial and equatorial conformations
of 1b–1e as model III at B3LYP/6-31G(d)
high axial conformation preference (Table 2, entry 4) but also
to the ‘‘deep’’ stepped conformation (Fig. 2) of the catalyst
and lack of sterical interactions of small ‘‘compact’’ diene with
axial diamine substituents. Such interactions probably
destabilize preferred transition states in two other model
reactions of the considerably more expansive reagents, leading
in consequence to lower enantioselectivities.
Entry
Cat.
R in diamine
DEa
28.2
DGa
31.8
1
2
3
4
1b
1c
1d
1e
Ph
Me
c-C6H11
t-Bu
ꢁ8.7
ꢁ5.5
ꢁ47.7
ꢁ45.9
ꢁ100.1
ꢁ101.5
a
In conclusion, the hypothesis that substituents on the
diamine moiety of the catalyst avoid the axial orientation
due to steric interaction with apical ligands has been revised.
Nevertheless, the stereochemical model with the new assumption
of catalyst conformation equilibrium has been shown to fit
the experimental data well. Moreover, we believe that the
conclusions that follow from the above observations also
have implications for epoxidation10 and other metallosalen-
catalyzed reactions.
The energy difference between the axial and equatorial conformation,
in kJ molꢁ1
.
reactivity of both sides of the carbonyl group was investigated
based on the example of the reaction with buta-1,3-diene.
The catalyst–formaldehyde complex derived from simple
ethylenediamine was set in one of the stepped conformations
and the transition states of the reaction on both enantiotopic
sides of aldehyde were calculated at the B3LYP/6-31G(d) level
of theory (Fig. 3). The approach of diene over the catalyst’s
downward phenyl ring, represented as a simple double bond in
model III, is preferred by 2.9 kJ molꢁ1. This approach has also
been preferred by 4.2 and 3.6 kJ molꢁ1 at the MP1WK/
6-31G(d) and M05/6-31G(d) levels of theory, respectively.
The calculation performed with the most sophisticated model
I provided similar results of 2.4 kJ molꢁ1 of energy difference
of the transition states.
Financial support from the Ministry of Science and
Higher Education (Grant PBZ-KBN-118/T09/16) is gratefully
acknowledged.
Notes and references
1 For review, see: E. M. McGarrigle and D. G. Gilheany,
Chem. Rev., 2005, 105, 1563.
2 For review, see: C. Schneider, Synthesis, 2006, 3919.
3 For review, see: P. G. Cozzi, Chem. Soc. Rev., 2004, 33, 410.
4 (a) For examples, see: Y. G. Abashkin and S. K. Burt, Org. Lett.,
2004, 6, 59; (b) L. Cavallo and H. Jacobsen, Angew. Chem., Int.
Ed., 2000, 39, 589; (c) J. El-Bahraoui, O. Wiest, D. Feichtinger and
D. A. Plattner, Angew. Chem., Int. Ed., 2001, 40, 2073;
(d) P. Fristrup, B. B. Dideriksen, D. Tanner and P.-O. Norrby,
J. Am. Chem. Soc., 2005, 127, 13672.
5 Y. Huang, T. Iwama and V. H. Rawal, J. Am. Chem. Soc., 2002,
124, 5950.
6 (a) P. Kwiatkowski, W. Cha"adaj, M. Malinowska,
M. Asztemborska and J. Jurczak, Tetrahedron: Asymmetry, 2005,
16, 2959; (b) P. Kwiatkowski, W. Cha"adaj and J. Jurczak,
Tetrahedron, 2006, 62, 5116.
7 (a) T. Hamada, T. Fukuda, H. Imanishi and T. Katsuki,
Tetrahedron, 1996, 52, 515; (b) Y. N. Ito and T. Katsuki,
Tetrahedron Lett., 1998, 39, 4325; (c) T. Hashihayata, Y. Ito and
T. Katsuki, Tetrahedron, 1997, 53, 9541.
8 For more information on the role of the conformation of metallo-
salen see: T. Katsuki, Adv. Synth. Catal., 2002, 344, 131.
9 (a) Y. Huang, T. Iwama and V. H. Rawal, J. Am. Chem. Soc.,
2000, 122, 7843; (b) Y. Huang, T. Iwama and V. H. Rawal,
Org. Lett., 2002, 4, 1163.
It is probable that both investigated transition states are
stabilized by C–Hꢂ ꢂ ꢂO interactions restricting the rotation of
aldehyde about the O–Cr bond. Such interactions creating an
additional organizing element in aldehyde–Lewis acid complex
has been proposed by Corey.15 Murphyet al. also considered
hydrogen bonds as organizing elements in epoxide–VO(salen)
complex.16
The calculated differences in the energy of transition states
correspond to the enantioselectivity of the reaction around
50–60% ee at room temperature, which is in agreement with
experimental data for catalyst 1a. Of course, enantioselectivity
depends also on the catalyst’s conformer population as well as
conformer geometry, which can be affected by the substituents
present in the structure of the catalyst. High enantioselectivity
of the 1e-catalyzed cycloadditions of cyclohexa-1,3-diene
(Table 1, entry 5), for instance, can be ascribed not only to
10 For example of the salen manganese catalyst violating Katsuki’s
model for epoxidation, see: A. Scheurer, H. Maid, F. Hampel,
R. W. Saalfrank, L. Toupet, P. Mosset, R. Puchta and N. J. R. van
Eikema Hommes, Eur. J. Org. Chem., 2005, 2566.
11 I. Iwakura, T. Ikeno and T. Yamada, Angew. Chem., Int. Ed.,
2005, 44, 2524.
12 S. E. Schaus, J. Branalt and E. N. Jacobsen, J. Org. Chem., 1998,
63, 403.
13 P. Kwiatkowski, M. Asztemborska, J.-C. Caille and J. Jurczak,
Adv. Synth. Catal., 2003, 345, 506.
14 D. J. Darensbourg, R. M. Mackiewicz, J. L. Rodgers, C. C. Fang,
D. R. Billodeaux and J. H. Reibenspies, Inorg. Chem., 2004, 43,
6024.
15 (a) E. J. Corey and T. W. Lee, Chem. Commun., 2001, 1321; (b) For
computational investigations see: M. N. Paddon-Row,
C. D. Anderson and K. N. Houk, J. Org. Chem., 2009, 74, 861,
and references cited therein.
16 D. M. Murphy, I. A. Fallis, D. J. Willock, J. Landon, E. Carter
and E. Vinck, Angew. Chem., Int. Ed., 2008, 47, 1414.
Fig. 3 Geometries of transition states leading to enantiomeric
products calculated at B3LYP/6-31G(d).
ꢀc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 6747–6749 | 6749