Key role of the Lewis base position in asymmetric bifunctional catalysis: Design and evaluation of a new ligand for chiral polymetallic catalysts

Article abstract of DOI:10.1021/ja067003h

New chiral ligands for asymmetric polymetallic catalysts were designed on the basis of the assumption that the higher-order assembly structure is stabilized by modifying the modular unit. The designed ligands 6 and 7 contained a scaffolding cyclohexane ring with a Lewis base phosphine oxide directly attached to the scaffold. A module in the polymetallic complex contains two metals per ligand, and a stable 6-, 5-, 5-membered fused chelation ring system should be generated. Synthesis of these ligands is simple and high yielding, using a catalytic dynamic kinetic resolution promoted by the Trost catalyst as a key step. Ligand function was assessed in a catalytic asymmetric ring-opening reaction of meso-aziridines with TMSCN, a useful reaction for the synthesis of optically active β-amino acids. The Gd complex generated from Gd(OⁱPr)₃ and the ligand was a highly active and enantioselective catalyst in this reaction. Enantioselectivity was reversed compared to the previously reported d-glucose-derived catalyst containing the same chirality of the individual module. ESI-MS analysis and X-ray crystallographic studies indicate that the assembly state of the modules in the polymetallic catalysts differs depending on the chiral ligand. The difference in the higher-order structure stems from a subtle change (one carbon) in the position of the Lewis base relative to the Gd metal. The change in the higher-order structure of the polymetallic complex led to a dramatic reversal of the enantioselectivity and increased catalyst activity. Copyright

Full text of DOI:10.1021/ja067003h

Published on Web 12/07/2006
Key Role of the Lewis Base Position in Asymmetric Bifunctional Catalysis:
Design and Evaluation of a New Ligand for Chiral Polymetallic Catalysts
Ikuo Fujimori,† Tsuyoshi Mita,† Keisuke Maki,† Motoo Shiro,‡ Akihiro Sato,§ Sanae Furusho,§
Motomu Kanai,*,† and Masakatsu Shibasaki*,†
Graduate School of Pharmaceutical Sciences, The UniVersity of Tokyo, Tokyo 113-0033, Japan,
Rigaku Corporation, 3-9-12 Matsubara, Akishima, Tokyo 196-8666, Japan, and
JASCO International Co., Ltd., 1-11-10 Myojin-Cho, Hachioji, Tokyo 192-0046, Japan
Received September 29, 2006; E-mail: mshibasa@mol.f.u-tokyo.ac.jp
Our group is involved in developing new asymmetric catalysts
based on the concept of bifunctional catalysis.1 Specifically, we
recently clarified the three-dimensional crystal structures of asym-
metric catalysts, useful in various cyanation reactions, containing
rare earth metals and chiral ligands derived from D-glucose (1 and
2).2 This previous study demonstrated that the active asymmetric
catalyst is a polymetallic complex, and that the higher-order
structure of the complex has profound effects on the function
(enantioselectivity and activity) of the asymmetric catalyst. In this
communication, we report that a subtle change of the chiral ligand
structure is amplified in the higher-order structure of the poly-
metallic complex, resulting in a dramatic difference in the function
of the asymmetric catalyst.
On the basis of previous crystal structures derived from ligands
Figure 1. Chiral ligands (1-4, 6, and 7) and schematic depiction of the
1 and 2,2 a catalytically active polymetallic complex was constructed
modular unit structures (5 and 8) in polymetallic complexes.
through self-assembly of the modular unit depicted as 5 (Figure
1). In module 5, the chiral ligand acts as a tetradentate ligand with
each ligand bridging two metals by forming a 7-, 5-, and
5-membered fused chelation ring system. Our hypothesis for the
design of a new asymmetric polymetallic catalyst was that its
higher-order structure will be more stable when assembled from
more stable modules. Thus, we designed new ligands 6 and 7 with
a truncated linker between the scaffolding cyclohexane ring and
the Lewis base phosphine oxide. In this case, module 8 contains a
6-, 5-, 5-membered fused ring system.
Scheme 1. Synthesis of Ligand 7
Synthesis of the designed ligands was simple and high yielding
using the catalytic hydrolytic dynamic kinetic resolution of allyl-
carbonate3 as the initial key step (Scheme 1).4 The thus-obtained
enantiomerically enriched allylic alcohol 10 was subjected to hy-
drogen-bonding-directed epoxidation, producing cis-epoxy alcohol
11 with excellent stereoselectivity (24:1). After introduction of the
catechol moiety through the Mitsunobu reaction, site-selective
epoxide opening with LiPPh2 proceeded concomitantly with methyl
ether cleavage at the phenolic oxygen. Enantiomerically pure 7 was
obtained after one recrystallization.
To identify the utility of new ligands 6 and 7, we focused on
the enantioselective ring-opening reaction of meso-aziridines with
TMSCN. We previously reported the first example of this reaction
type using a polymetallic Gd catalyst (2:3 or 4:5+oxo complex of
Gd and 3) generated from Gd(OiPr)3 and 3 in a 1:2 ratio.5 Although
the substrate generality was broad, the catalyst activity and
enantioselectivity were not completely satisfactory (see Table 1,
entry 3 for a typical result). Because this type of reaction is useful
for the synthesis of enantiomerically enriched â-amino acidss
important chiral building blocks for â-peptides and foldamers6sa
more efficient asymmetric catalyst was desirable.
We began our studies by applying the previously optimized
conditions (Table 1, entry 3) to a ring-opening reaction of aziridine
13a using ligand 7. As a result, amido nitrile 14a was obtained
with 52% ee (99% yield in 15 h at room temperature). Through
optimization of the reaction conditions, we identified several
important deviations from the previously observed tendency: (1)
Addition of a catalytic amount of TFA produced detrimental effects
using 7. (2) Reaction in THF produced higher enantioselectivity
than in propionitrile. (3) The enantioselectivity was consistently
high (>95% ee) regardless of the Gd/ligand ratio (from 1:1 to 1:4)
in the catalyst preparation.7 These dependencies are in sharp contrast
to the previous results using ligand 3, in which the enantioselectivity
was strongly dependent on the Gd/ligand ratio. (4) Although 7
contains the same chirality as 3, the product produced by the
catalysts deriVed from these ligands had reVersed enantioselectiVity.
Under the optimized conditions, we next investigated the scope
of this catalysis (Table 1). Excellent enantioselectivity was produced
† The University of Tokyo.
‡ Rigaku Corporation.
§ JASCO International Co., Ltd.
9
16438
J. AM. CHEM. SOC. 2006, 128, 16438-16439
10.1021/ja067003h CCC: $33.50 © 2006 American Chemical Society

C O M M U N I C A T I O N S
Table 1. Catalytic Enantioselective Ring-Opening Reaction of
meso-Aziridines with TMSCN
Figure 2. X-ray structure of 3:4+2OH complex (a) and its chemical
depiction (b).
from a catalyst solution in propionitrile. Crystallographic analysis
revealed that the crystal was a Gd/ligand ) 3:4+2OH complex
(Figure 2).7 Using this crystal as a precatalyst (Gd ) 12.5 mol %),
product 14a was obtained with only 62% ee. On the basis of the
ESI-MS observations, the composition of 3:4+2OH was maintained
when the crystal was dissolved in solution. Thus, the crystal
structure does not represent the actual catalyst of the aziridine
opening reaction. The enantioselectivity, however, recovered to 94%
ee when Gd(OiPr)3 was added to the crystal solution to adjust the
Gd/ligand ratio to 5:6 and heated at 50 °C for 1 h before the reaction
was conducted. Therefore, the assembly-state conversion from the
5:6+oxo+OH complex to the 3:4+2OH complex in the crystal-
lization process is reversible under appropriate conditions. Impor-
tantly, the crystal structure in Figure 2 strongly supports our initial
hypothesis that the module contains a stable fused tricyclic chelation
structure, in which one ligand binds to two Gd metals.
In summary, we identified new chiral ligands (6 and 7) for
asymmetric polymetallic catalysts. The catalysts demonstrated
improved, reVersed enantioselectivity as well as significantly higher
activity in an asymmetric ring-opening reaction of meso-aziridines
with TMSCN, compared to the previous D-glucose-derived catalysts.
The functional difference between the two catalyst groups is
attributable to a higher-order structure change caused by a subtle
modification in the position of the Lewis base. Further studies to
elucidate the active catalyst structure and extension of the current
concept to other reactions are ongoing.
b
a Isolated yield. Determined by chiral HPLC. c In the presence of 10
mol % of 2,6-dimethylphenol at room temperature. d Using a catalyst
generated from 10 mol % of Gd(OiPr)3 and 20 mol % of 3 in the presence
of 5 mol % of TFA and 1 equiv of 2,6-dimethylphenol in propionitrile at
0 °C. Product with the opposite configuration (S,S) was obtained. See ref
5. e Absolute configuration was determined as shown (R,R). f Reaction
temperature ) 60 °C.
from a range of aziridines using 2 mol % of the catalyst. When
using 13a as a substrate, catechol-containing ligand 6 produced
better results than 7 (entry 2). Thus, significantly higher enantio-
selectivity and catalyst turnover8 were generally produced using
catalysts prepared from 6 or 7 than when using the previous
D-glucose-derived ligand 3.
Comparing the two asymmetric catalysts prepared from D-
glucose-derived ligands (1-3) and the new ligands (6 and 7), the
contrasting dependency of the enantioselectivity on the Gd/ligand
ratio in catalyst preparation and the reverse enantioselectivity
suggest that the modular assembly state in the polymetallic catalysts
differs depending on the chiral ligands. This assumption was
supported by ESI-QFT-MS observation; while the Gd/ligand ) 2:3
and 4:5+oxo complexes were the main species when the catalyst
was prepared from Gd(OiPr)3 and 1-3 in a 1:2 ratio,2 the
5:6+oxo+OH complex [MW ) 3474.3881 (M + H)+] was the
sole species in a catalyst solution prepared from Gd(OiPr)3 and 7
in both 1:1.2 and 1:4 ratios.7
The dramatic change in the modular assembly state of the
polymetallic complexes was not due to the absence of the oxygen
atom in the scaffolding cyclohexane ring in 6 and 7; using a catalyst
prepared from Gd(OiPr)3 and a control ligand 4 in a 1:2 ratio (10
mol %), we obtained 14a with the same absolute configuration to
that produced by D-glucose-derived ligands and with 44% ee.
Therefore, the higher-order structure change of the polymetallic
complex and the resulting change in the function of the asymmetric
catalyst are likely due to the one carbon difference in the relative
position of the phosphine oxide and Gd.
Acknowledgment. Financial support was provided by a Grand-
in-Aid for Specially Promoted Research of MEXT.
Supporting Information Available: Catalyst preparation, spectra,
and cyrstallographic data (CIF). This material is available free of charge
via the Internet at http://pubs.acs.org.
References
(1) (a) Shibasaki, M.; Sasai, H.; Arai, T. Angew. Chem., Int. Ed. Engl. 1997,
36, 1236. (b) Kanai, M.; Kato, N.; Ichikawa, E.; Shibasaki, M. Synlett
2005, 1491.
(2) Kato, N.; Mita, T.; Kanai, M.; Therrien, B.; Kawano, M.; Yamaguchi,
K.; Danjo, H.; Sei, Y.; Sato, A.; Furusho, S.; Shibasaki, M. J. Am. Chem.
Soc. 2006, 128, 6768.
(3) (a) Lu¨ssem, B. J.; Gais, H.-J. J. Am. Chem. Soc. 2003, 125, 6066. (b)
Trost, B. M.; Organ, M. G. J. Am. Chem. Soc. 1994, 116, 10320.
(4) On the other hand, synthesis of 3 requires eight steps (overall yield )
37%) from a commercially available compound. See, Kato, N.; Tomita,
D.; Maki, K.; Kanai, M.; Shibasaki M. J. Org. Chem. 2004, 69, 6128.
(5) Mita, T.; Fujimori, I.; Wada, R.; Wen, J.; Kanai, M.; Shibasaki, M. J.
Am. Chem. Soc. 2005, 127, 11252.
(6) For reviews, see: (a) Abele, S.; Seebach, D. Eur. J. Org. Chem. 2000, 1.
(b) Gellman, S. H. Acc. Chem. Res. 1998, 31, 173.
(7) See Supporting Information for details.
(8) For example, the reaction of 13a was completed at room temperature in
4 h and in 15 min using 10 mol % of catalyst derived from 3 and 7,
respectively.
Studies toward elucidation of the three-dimensional structure of
the active catalyst derived from 7 led us to isolate colorless prisms
JA067003H
9
J. AM. CHEM. SOC. VOL. 128, NO. 51, 2006 16439