Structure-based analysis and optimization of a highly enantioselective catalyst for the strecker reaction

Article abstract of DOI:10.1021/ja027246j

A mechanistic investigation of the asymmetric Strecker reaction catalyzed by a metal-free Schiff base catalyst was conducted. The active site of the catalyst, the relevant stereoisomer of the imine substrate, and the solution structure of the imine-catalyst complex were elucidated using a series of kinetics, structure-activity, and NMR experiments. An unusual bridging interaction between the imine and the urea hydrogens of the catalyst was identified and supported by computation. Rational optimization of catalyst structure based on the mechanistic insight led to an improved catalyst for the asymmetric Strecker reaction. Copyright

Full text of DOI:10.1021/ja027246j

Published on Web 08/03/2002
Structure-Based Analysis and Optimization of a Highly Enantioselective
Catalyst for the Strecker Reaction
Petr Vachal and Eric N. Jacobsen*
HarVard UniVersity, Department of Chemistry and Chemical Biology, Cambridge, Massachusetts 02138
Received June 10, 2002
Scheme 1. Asymmetric Strecker Reaction Catalyzed by 1
The hydrocyanation of imines (the Strecker reaction) has become
one of the most intensively studied reactions in the field of
asymmetric catalysis over the past several years.1 A particularly
intriguing system identified in the course of these investigations is
the structurally novel metal-free catalyst 1, which displays high
enantioselectivity with extraordinary substrate scope in the hydro-
cyanation of aldimines and methylketoimines (Scheme 1).2 In a
compelling illustration of the potential role of combinatorial
strategies for the development of new selective catalyst systems, 1
was discovered and optimized by a parallel library synthesis and
screening approach without any direct understanding of how it
functions. At this stage, however, gleaning insight into the mode
of action of this catalyst becomes important both fundamentally
(given its remarkable effectiveness) and practically (since it is likely
that the full scope and utility of this class of catalysts cannot be
elucidated without a solid mechanistic foundation). We report here
structural and mechanistic studies that shed light on this issue and
lay the foundation for rational catalyst optimization.
Despite its relatively small size (fw ) 621 g/mol), catalyst 1
was found to adopt a well-defined secondary structure in solution.
The ground state conformation was determined through ROESY
and NOE NMR experiments in d8-THF and d8-dioxane solutions
of 1 (Figure 1A)3 and found to be in full accord with the calculated
energy-minimized geometry.4 The 3-D structure of 1 in solution
was independent of the solvent used for the NMR study,5 and
investigation of the NMR solution structures of several related
analogues of 1 revealed closely similar geometries.6
In an effort to ascertain the precise role of 1 in the Strecker
reaction, we carried out rate studies of the hydrocyanation of a
model substrate (N-allyl-4-methoxybenzaldimine). The transforma-
tion was found to obey Michaelis-Menten kinetics, with a first
order dependence on catalyst and HCN, and saturation kinetics with
respect to the imine substrate (KM ) 0.214 ( 0.009 M, kcat/KM
)
3.8 × 10-3 M-1 s-1). This result implicates reversible formation
of an imine-catalyst complex, presumably through a hydrogen bond
between the imine nitrogen and an acidic proton of the catalyst.
To identify the relevant proton(s), we prepared a series of analogues
of 1 and evaluated them as catalysts for activity and enantioselec-
tivity in the Strecker reaction.7 Remarkably, only the two urea
hydrogens of 1 were found to be essential for catalyst activity.
Isotope shift experiments provided further unequivocal evidence
that the imine substrate interacts solely with the urea hydrogens.8
The N-alkylimine substrates utilized by 1 exist as directly
observable mixtures of Z- and E-stereoisomers that interconvert
rapidly in solution.10 To develop a useful understanding of the nature
of the interaction between imines and 1, it was crucial to determine
which imine stereoisomer is involved in the substrate-catalyst
complex. NMR titration of a solution of a representative ketoimine
Figure 1. (A) Solution structure of catalyst 1 and (B, C) two views of the
complex generated upon binding of a Z-imine, as determined by NMR
analysis (see text). The images were generated using MOLMOL.10 The imine
employed in the NMR studies was N-p-methoxybenzyl acetophenone imine
(R ) p-methoxybenzyl). For clarity, the N-substituent R in the graphic is
represented as an allyl group. The imine is shown hydrogen-bonded to both
urea hydrogens in agreement with both experimental and computational
data (see text).
derivative (N-p-methoxybenzyl acetophenone imine; E:Z ) 20:1)
with catalyst 1 resulted in a downfield shift of the Z-imine methyl
resonance exclusively. Furthermore, a cyclic Z-imine, 3,4-dihy-
droisoquinoline, was found to be a highly competent substrate for
the Strecker reaction, undergoing hydrocyanation in quantitative
yield and in 89% ee.11 The sense of absolute stereoinduction was
identical to that of all other Strecker adducts derived from acyclic
imines that exist predominantly as E-isomers. Taken together, these
* Address correspondence to this author. E-mail: jacobsen@chemistry.harvard.edu.
9
10012
J. AM. CHEM. SOC. 2002, 124, 10012-10014
10.1021/ja027246j CCC: $22.00 © 2002 American Chemical Society

C O M M U N I C A T I O N S
Table 1. Model-Driven Optimization of Catalyst 1
Figure 2. Calculated geometry (energy minimum) of a catalyst-imine
complex (A) and a catalyst-product complex (B). All calculations were
performed on a simplified relevant system: N,N′-dimethylurea or N,N′-
dimethylthiourea, (Z)-ethylidene methylamine, and 2-(methylamino)propi-
onitrile using Gaussian 98 (B3LYP, 6-31G(d,p) basis set.
catalyst
eea (%)
catalyst
eea (%)
data provide compelling evidence that Strecker reactions with 1
involve binding of the imine substrate as the Z-isomer.
1
2
3
80.0
93.5
93.1
4
5
6
95.8
96.6
97.0
With this insight into the general features of the imine-catalyst
interaction in hand, we sought to establish a detailed 3-D structure
of the substrate-bound complex. Multiple NOE interactions between
1 and a variety of Z-imines were observable, allowing determination
of the orientation of these substrates relative to the catalyst.6
Intramolecular cross-peaks due to catalyst in the bound and free
states were essentially invariant, indicating that no significant
change in conformation of 1 results from binding of the imine.12
Depictions of the structure of the complex between 1 and a Z-imine
substrate fully consistent with all of the NOE data are shown in
Figure 1B,C. The imine substrate is placed in a bridging mode
between the two urea hydrogens. This structural assignment is in
line with the experimental data13 and supported by high level
calculations (Gaussian 98, B3LYP level with the 6-31G(d,p) basis
set)4 on a simplified model of the bound complex. Energy
minimization revealed a significant preference for a bridged
structure, with imine hydrogen-bonded to both urea hydrogens
simultaneously (Figure 2A).14 In contrast, analogous calculations
assign a singly hydrogen-bonded structure to the product amino-
nitrile-catalyst complex (Figure 2B). This establishes a plausible
explanation for the basis for catalyst turnover: the bridging
interaction in the catalyst-imine complex is stronger (8.5 kcal/
mol for urea; 10.0 kcal/mol for thiourea) than the classical hydrogen
bond in catalyst-product complex (5.0 and 6.3 kcal/mol, respec-
tively).15
The structure elucidated for the catalyst-substrate complex sheds
substantial light on the basis for the scope and selectivity of
asymmetric Strecker reactions with 1: (1) The large group on the
imine carbon is directed away from the catalyst and into solvent
(Figure 1B); this serves to explain why 1 catalyzes hydrocyanation
of most aldimines with high ee, regardless of the steric and
electronic properties of the substrate. (2) The small group (H for
aldimines, Me for methylketoimines) is aimed directly into the
catalyst; ketoimines bearing larger substituents are poor substrates
for the reaction,2c presumably because they cannot be accom-
modated within the optimal geometry. (3) The N-substituent is also
directed away from the catalyst. However, its size is restricted as
a result of the requirement to access the Z-isomer of the imine. (4)
On the basis of the observed sense of stereoinduction, addition of
HCN takes place over the diaminocyclohexane portion of the
catalyst (i.e., from the right-hand side in Figure 1C) and away from
the amino acid/amide portion.
a Determined by chiral GC analysis on γ-TA column.
Table 2. Comparison of Asymmetric Strecker Reactions of
Selected Aldimines and Ketoimines Using Original Catalyst 1 and
Optimized Catalyst 6
substrate
eea of product (%)
catalyst 1 catalyst 6
1
2
3
entry
R
R
R
1
2
3
4
5
6
i-Pr
n-Pent
t-Bu
Ph
t-Bu
Ph
H
H
Me
Me
H
Ph
Ph
Ph
p-BrC6H4
Ph
Ph
80
79
70
92
96
96
97
96
86
96
99.3
99.3
H
a Determined by GC and HPLC analyses using commercial chiral
columns.
2-6 and tested them as catalysts in the asymmetric Strecker reaction
of a substrate that performed relatively poorly with catalyst 1 (80%
ee, Table 1). Replacement of the secondary amide of 1 with a
bulkier tertiary amide (2-4) led to significant improvements in
enantioselectivity. Likewise, increasing the steric bulk of the amino
acid side chain had a beneficial effect (compare 4 vs 5). Given the
critical role identified for the urea in substrate binding, we
anticipated that the reaction outcome would be highly sensitive to
fine tuning of its electronic properties. Indeed, replacement of the
urea with a thiourea group led to a measurable improvement in
enantioselectivity (compare 4 vs 6). Thus, through this mechanism-
driven optimization exercise, catalyst 6 was identified as the most
enantioselective Strecker catalyst prepared to date.
The benefits of catalyst 6 relative to 1 extend over a wide range
of substrates. Imines that had represented limitations for this
Strecker methodology now undergo hydrocyanation with syntheti-
cally useful ee’s (Table 2, entries 1-3). Substrates that performed
relatively well with 1 can now be transformed with nearly perfect
enantiocontrol (entries 5 and 6).
A synthetically useful chiral catalyst that had been developed
by a purely empirical approach has evolved into a system amenable
to rational design and improved significantly on the basis of detailed
mechanistic and structural insights. The very unusual basis for
substrate activation emerges as one of the most intriguing findings
of this study. Further work will be directed toward elucidating the
The last hypothesis leads to the prediction that increasing the
steric properties of the amino acid/amide portion of the catalyst
should lead to higher enantioselectivity in the hydrocyanation
reactions. To test this notion, we prepared a series of derivatives
9
J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002 10013

C O M M U N I C A T I O N S
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; Johnson, B. G.; Chen,
W.; Wong, M. W.; Andres, J. L.; Head-Gordon, M.; Replogle, E. S.; Pople,
J. A. Gaussian 98, revision 5.4; Gaussian, Inc.: Pittsburgh, PA, 1998.
(5) NMR data obtained in THF and dioxane were indistinguishable. Hydro-
cyanations catalyzed by 1 afford essentially identical enantioselectivity
in a wide range of nonprotic solvents (toluene, benzene, THF, dioxane,
or hexanes).
details of the HCN addition step and toward evaluating the potential
of this catalyst system for other enantioselective reactions of imines.
Acknowledgment. This work was supported by the NIH (GM-
43214) and through fellowship support to P.V. from A. Bader,
BMS, and Eli Lilly. The authors thank Dr. T. Kuribayashi, Dr. M.
S. Sigman, and A. G. Wenzel for important background experiments
and Q. Chen, Dr. K. Gademann, Dr. S. Huang, Dr. Z. Yu, and R.
Ruck for valuable discussions.
(6) Full details provided as Supporting Information.
(7) Deletion of the secondary amide proton, the phenolic proton, and the imine
functionality did not suppress catalytic activity. In contrast, alkylation of
either of the urea nitrogens or replacement with a carbamate group led to
dramatic loss of activity and enantioselectivity. Details are provided as
Supporting Information.
Supporting Information Available: Experimental procedures and
details of the structural analyses (PDF). This material is available free
of charge via the Internet at http://pubs.acs.org.
(8) A solution of the catalyst in d8-THF was split evenly into two NMR tubes;
N-benzylpivalaldimine (N-(2,2-dimethylprpylidene)benzylamine) was added
to the first NMR tube; to the second an identical quantity of the
corresponding 15N-labled imine was added. 1H NMR spectra were recorded
and were found to be completely superimposable (∆ δ for all resonances
) <1 Hz) with the exception of both urea hydrogens (∆ δ ) 16.1 and
16.9 Hz). Direct through-bond coupling (N-H or N-H-N) was not
observed, but this was expected given the noncovalent and rapidly
exchanging nature of the interaction. For selected examples of studies
involving chemical shift invoked by isotope change, see: (a) Benedict,
H.; Hoelger, C.; Aquilar-Parrilla, F.; Fehlhammer, P.; Wehlan, M.;
Janoschek, R.; Limbach, H. H. J. Mol. Struct. 1996, 378, 11. (b) Hansen,
P. E.; Hansen, A. E.; Lycka, A.; Buvari-Barcza, A. Acta Chim. Scand.
1993, 47, 777. (c) Jameson, C. J.; Jameson, A. J.; Oppusunggu, D. J.
Chem. Phys. 1986, 85, 5480.
References
(1) (a) Iyer, M. S.; Gigstad, K. M.; Namdev, N. D.; Lipton, M. J. Am. Chem.
Soc. 1996, 118, 4910. (b) Iyer, M. S.; Gigstad, K. M.; Namdev, N. D.;
Lipton, M. Amino Acids 1996, 11, 259. (c) Sigman M. S.; Jacobsen, E.
N. J. Am. Chem. Soc. 1998, 120, 5315. (d) Ishitani, H.; Komyiama, S.;
Kobayashi, S. Angew. Chem., Int. Ed. 1998, 37, 3186. (e) Corey, E. J.;
Grogan, M. J. Org. Lett. 1999, 1, 157. (f) Krueger, C. A.; Kuntz, K. W.;
Dzierba, C. D.; Wirschun, W. G.; Gleason, J. D.; Snapper, M. L.; Hoveyda,
A. H. J. Am. Chem. Soc. 1999, 121, 4284. (g) Porter, J. R.; Wirschun, W.
G.; Kuntz, K. W.; Snapper, M. L.; Hoveyda, A. H. J. Am. Chem. Soc.
2000, 122, 2657. (h) Kobayashi, S.; Ishitani, H. Chirality 2000, 12, 540.
(i) Takamura, M.; Hamashima, Y.; Yoshitaka, U.; Usuda, H.; Kanai, M.;
Shibisaki, M. Angew. Chem., Int. Ed. 2000, 39, 1650. (j) Byrne, J. J.;
Chavarot, M.; Chavant, P.-Y.; Vallee Y. Tetrahedron Lett. 2000, 41, 873.
(k) Ishitani, H.; Komyiama, S.; Hasegawa, Y.; Kobayashi, S. J. Am. Chem.
Soc. 2000, 122, 762. (l) Takamura, M.; Funabashi, K.; Kanai, M.;
Shibasaki, M. J. Am. Chem. Soc. 2000, 122, 6327. (m) Takamura, M.;
Hamashima, Y.; Usuda, H.; Kanai, M.; Shibasaki, M. Chem. Pharm. Bull.
2000, 48, 1586. (n) Kanai, M. Yakugaku Zasshi 2001, 121, 949. (o)
Josephsohn, N. S.; Kuntz, K. W.; Snapper, M. L.; Hoveyda, A. H. J. Am.
Chem. Soc. 2001, 123, 11594. (p) Kanai, M.; Hamashima, Y.; Takamura,
M.; Shibasaki, M. Yuki Gosei Kagaku Kyokaishi 2001, 59, 766. (q)
Chavarot, M.; Byrne, J. J.; Chavant, P. Y.; Vallee, Y. Tetrahedron:
Asymmetry 2001, 12, 1147. (r) Nogami, H.; Matsunaga, S.; Kanai, M.;
Shibasaki, M. Tetrahedron Lett. 2001, 42, 279. (s) Takamura, M.;
Funabashi, K.; Kanai, M.; Shibasaki, M. J. Am. Chem. Soc. 2001, 123,
6801. (t) Funabashi, K.; Ratni, H.; Kanai, M.; Shibasaki, M. J. Am. Chem.
Soc. 2001, 123, 10784. For a recent overview, see: (u) Yet, L. Angew.
Chem., Int. Ed. 2001, 40, 875.
(9) Koradi, R.; Billeter, M.; Wu¨thrich, K. J. Mol. Graphics Modell. 1996,
14, 51.
(10) Irradiation of resonances due to either the E- or Z-imine stereoisomers
results in observable saturation of the corresponding resonances of the
other isomer. For theoretical details on energy transfer experiments invoked
by chemical exchange, see: (a) Sanders, J. K. M.; Hunter, B. K. Modern
NMR Spectroscopy A Guide for Chemists; Oxford University Press: New
York, 1989. (b) Alger, J. R.; Shulman, R. G. Quart. ReV. Biophys. 1984,
17, 83. (c) Sanders, J. K. M.; Merch, J. D. Prog. Nucl. Magn. Reson.
Spectrosc. 1982, 15, 353. (d) Boyd, J.; Brindle, K. M.; Campbell, I. D.;
Radda, G. K. J. Magn. Reson. 1984, 60, 149.
(11) In contrast, cyclic imines restricted to E-configurations (e.g., 6-phenyl-
2,3,4,5-tetrahydropyridine) underwent no reaction under the same condi-
tions, even under extended reaction times.
(12) On the basis of KM values determined in the kinetics studies, it was
calculated that >80% of 1 was bound to substrate under the conditions
of the NMR experiments.
(13) In addition to NOE cross-peak data, which are consistent with the
placement of imine in such an orientation, titration of solutions of 1 with
Z-imines results in upfield shifts of the two urea protons to similar degrees.
This can be ascribed either to a rapidly equilibrating structure where the
imine shifts back and forth between the two urea protons or to a static,
bridged structure. A strong preference for the latter is suggested by the
computational studies. It must be noted, however, that such a bridging
structure may not be maintained in the intermediates and transition states
generated upon addition of HCN to the medium.
(2) (a) Sigman M. S.; Jacobsen, E. N. J. Am. Chem. Soc. 1998, 120, 4901.
(b) Sigman, M. S.; Vachal, P.; Jacobsen, E. N. Angew. Chem., Int. Ed.
2000, 39, 1279. (c) Vachal, P.; Jacobsen, E. N. Org. Lett. 2000, 2, 867.
(d) Su, J. T.; Vachal, P.; Jacobsen, E. N. AdV. Synth. Catal. 2001, 343,
197.
(3) The values of the intramolecular NOE cross-peak volumes and the
calculated distances are provided as Supporting Information.
(4) Energy minimization was performed in Gaussian at the B3LYP level using
the 3-21G* basis set. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.;
Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.;
Montgomery, J. A., Jr.; Stratmann, R. E.; Burant, J. C.; Dapprich, S.;
Millam, J. M.; Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.;
Tomasi, J.; Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli,
C.; Adamo, C.; Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.;
Cui, Q.; Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari,
K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;
(14) For examples of X-ray crystal structures in which analogous interactions
are observed, see: (a) Carpy, P. A.; Leger, J.-M.; Wermuth, C.-G.; Leclerc,
G. Acta Crystallogr., Sect. B 1981, 37, 885. (b) Lock, C. J. L.; Pilon, P.;
Lippert, B. Acta Crystallogr., Sect. B 1979, 35, 2533.
(15) All hydrogen bond energies as well as energy minimizations were
calculated for the gas phase. Given the absence of solvent dependence in
the asymmetric Strecker reaction noted in ref 5, this significant simplifica-
tion appears justified.
JA027246J
9
10014 J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002