Cyclopropenimine-catalyzed enantioselective mannich reactions of tert -butyl glycinates with N -Boc-imines

Article abstract of DOI:10.1021/ja407277a

Cyclopropenimine 1 is shown to catalyze Mannich reactions between glycine imines and N-Boc-aldimines with high levels of enantio- and diastereocontrol. The reactivity of 1 is shown to be substantially greater than that of a widely used thiourea cinchona alkaloid-derived catalyst. A variety of aryl and aliphatic N-Boc-aldimines are effective substrates for this transformation. A preparative-scale reaction to deliver >90 mmol of product is shown using 1 mol % catalyst. The products of this transformation can be converted into several useful derivatives.

Full text of DOI:10.1021/ja407277a

Communication
pubs.acs.org/JACS
Cyclopropenimine-Catalyzed Enantioselective Mannich Reactions of
tert-Butyl Glycinates with N‑Boc-Imines
Jeffrey S. Bandar and Tristan H. Lambert*
Department of Chemistry, Columbia University, New York, New York 10027, United States
S
* Supporting Information
employed N-sulfonyl imine electrophiles, which result in
products that are challenging to deprotect.7,8 More synthetically
ABSTRACT: Cyclopropenimine 1 is shown to catalyze
Mannich reactions between glycine imines and N-Boc-
aldimines with high levels of enantio- and diastereocontrol.
The reactivity of 1 is shown to be substantially greater than
that of a widely used thiourea cinchona alkaloid-derived
catalyst. A variety of aryl and aliphatic N-Boc-aldimines are
effective substrates for this transformation. A preparative-
scale reaction to deliver >90 mmol of product is shown
using 1 mol % catalyst. The products of this trans-
formation can be converted into several useful derivatives.
versatile products have been obtained with chiral phase-transfer
catalysis,9 but typically with only moderate levels of stereo-
selectivity.10 Arguably the most attractive approach to this class
of reactions is via direct chiral Brønsted-base catalysis,11 which
has been shown to be compatible with the use of N-carbamoyl
imine substrates. Indeed, the state-of-the-art catalyst for this
transformation is the widely employed bifunctional thiourea 5,12
which relies on strong H-bond activation coupled with a
relatively weak tertiary amine Brønsted base (eq 1). Barbas has
icinal diamino stereoarrays represent a prominent chemical
V
motif found in bioactive natural products, synthetic
building blocks, and metal ligands.1 The preparation of α,β-
diamino acid derivatives via enantioselective direct Mannich
reactions of glycinate Schiff bases offers a powerful means to
access such stereoarrays (Figure 1).2 Although a number of
Figure 1. Vincinal diamino stereocenters via direct glycinate Mannich
reaction.
reported that 5 catalyzes the stereoselective Mannich reaction of
methyl glycinate substrates in 24−36 h but is ineffective for the
more sterically demanding tert-butyl esters, which are desirable
because of their amenability to acidic deprotection.13 Notably,
the more strongly basic catalyst tetramethylguanidine (TMG) is
also ineffective at catalyzing the reaction shown in eq 1 (R =
Me).14
Given the remarkably high reactivity that we had previously
observed for Michael additions using cyclopropenimine 1,3 we
wondered whether this catalyst could also offer a more broadly
effective approach to glycinate Mannich reactions and thereby
help to alleviate the substrate limitations noted above. In fact, we
found that 10 mol % 1 catalyzed the reaction of glycinate 3 (R =
Me) with N-Boc-benzaldimine (2) to produce diamine adduct 4
in 81% yield with 95:5 dr (syn/anti) and 95% ee in only 15 min at
enantioselective catalyst systems that deliver α-amino Mannich
products efficiently with high stereoselectivity have been
reported, these methods suffer from important limitations of
substrate scope, catalyst reactivity, or stereocontrol, and thus, a
more broadly effective solution to the challenge of enantiose-
lective Mannich reactions remains to be developed. We recently
introduced chiral cyclopropenimines as a potent new class of
enantioselective Brønsted base catalysts.3 Here we report that the
readily available4,5 chiral cyclopropenimine 1 is a highly efficient
and selective catalyst for direct Mannich reactions that operates
even in the context of the challenging coupling of tert-butyl
glycinate donors and N-Boc-imine acceptors.
Enantio- and diastereoselective Mannich reactions of glycinate
Schiff bases6 have been accomplished by a number of groups
using copper-based catalysis; however, these methods have
Received: July 16, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja407277a | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
room temperature, a remarkable enhancement of reactivity over
thiourea catalyst 5. Moreover, 1 was also found to catalyze the
reaction of tert-butyl glycinate 3 in 20 h to deliver the Mannich
product in 97% yield with 99:1 dr and 94% ee.15 These results
clearly demonstrate the substantially greater reactivity of 1 versus
less basic catalyst frameworks.16
Table 2. Gram-Scale Substrate Scope of the Mannich
Reaction
a
A portion of the optimization studies leading to the reaction
shown in eq 1 are presented in Table 1. In all cases, 4 Å molecular
Table 1. Optimization of Enantioselective Cyclopropenimine-
a
Catalyzed Mannich Reaction
entry
solvent
conc (M)
time (h)
syn/anti
ee (%)
1
2
3
4
5
6
7
8
9
EtOAc
Et2O
0.35
0.35
0.35
0.35
0.35
0.45
0.25
0.15
0.08
0.25
0.25
12
8
98:2
98:2
99:1
99:1
99:1
99:1
99:1
99:1
98:2
98:2
98:2
83
85
89
91
91
90
92
94
93
95
95
PhH
20
20
16
6
m-xylene
PhMe
PhMe
PhMe
PhMe
PhMe
PhMe
PhMe
24
36
48
20
26
b
10
bc
,
11
a
All reactions proceeded to ≥95% conversion of glycine imine.
b
c
Ground 4 Å molecular sieves were used. 1.5 equiv of 2 was used.
sieves were added to inhibit hydrolysis of imine 2, which 1 slowly
catalyzes in the presence of water. Although the use of solvents
such as ethyl acetate and diethyl ether resulted in promising
selectivities (entries 1 and 2), we found that the enantiose-
lectivity was increased in aromatic solvents (entries 3−5), with
toluene providing optimal results. Dilution of the reaction
mixture increased the selectivity, albeit at a cost of reaction time
(entries 5−9). Additionally, we found that using ground
molecular sieves improved the efficiency and selectivity of the
reaction (entry 10). Notably, the stoichiometry of the electro-
phile could be reduced to 1.5 equiv with no loss in efficiency or
selectivity (entry 11).
Next, the substrate scope of the cyclopropenimine-catalyzed
Mannich reaction was explored (Table 2). To emphasize the
practicability and simplicity of this chemistry, each of the
reactions shown was performed using 1 g of glycinate substrate
under conditions that included exposure to the atmosphere.
Methyl, benzyl, and tert-butyl glycinates all showed good
reactivity and selectivity at this scale with N-Boc-benzaldimine
(entry 1). Given the incompatibility of tert-butyl glycinates with
many other Brønsted base catalysts, we chose to explore the
reaction of this substrate with other N-Boc-aldimines. Both o-
and p-tolualdimines yielded diamine products with high
diastereo- and enantioselectivity (entries 2 and 3). The reaction
with the more electron-rich anisaldimine was also efficient and
stereoselective (entry 4), although the reaction time of 60 h was
significantly longer than that with electron-neutral substrates. In
this case, use of the methyl glycinate substrate led to product in
only 1.5 h with comparably good stereoselectivity. Halogenated
a
Yields based on purified products. Diastereomeric ratios (dr) and
b
enantiomeric excesses (ee) were determined by HPLC. Yield
determined by 1H NMR versus Bn2O as a standard; product
c
characterized after hydrolysis of the benzophenone imine. Reaction
was performed at a concentration of 0.07 M. 20 mol % catalyst was
used; 0.9 mmol scale.
d
substrates were found to perform well (entries 5 and 6); however
the highly electron-deficient p-trifluoromethylbenzaldimine was
relatively unreactive and yielded the diamine with severely
diminished ee (entry 7). Notably, heterocyclic imines, including
those with pyridine, furan, and thiophene ring systems, were all
excellent substrates for this transformation (entries 8−10).
Interestingly, although the enantioselectivity with the pyridyl
substrate was subpar using the tert-butyl glycinate donor, high
selectivity was restored with the use of the methyl glycinate
(entry 8). An alanine-derived imine substrate gave rise to a
product bearing a tetrasubstituted stereocenter (entry 11);
however, the diastero- and enantioselectivity of this substrate
type were suboptimal with the current catalyst system.
B
dx.doi.org/10.1021/ja407277a | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
Of particular note is the fact that an aliphatic aldimine, 6, was
also found to be a suitable substrate for reaction with the tert-
butyl glycinate donor 3b (eq 2). Although this reaction required
Substrates bearing isopropyl and ethyl groups resulted in high
diastereo- and enantioselectivities (entries 1 and 2). Entry 3
shows the production of protected 2,3-diaminobutanoic acid, an
α,β-diamino acid widely found as a key motif in a number of
naturally occurring peptides.1 For this substrate, the N-Boc-
aldimine was formed in situ because of its instability and difficulty
of isolation. At 0 °C, the reaction proceeded with good
conversion, albeit with only 44% ee. The selectivity could be
increased to 83% ee upon cooling to −78 °C, but at this
temperature the reaction did not reach completion.17 We also
found silyl-protected alcohol (entry 4) and N-Boc-protected
amine (entry 5) functional groups to be well-accommodated.
Additionally, a substrate bearing a terminal olefin worked well
with high selectivities (entry 6). The tolerance of diverse
functional groups in this reaction should enable an enhanced
synthetic utilization of these vicinal diamino stereoarrays.
a 20 mol % catalyst loading to achieve a reasonable rate, the
procedure was readily performed on a gram scale to produce
adduct 7 with high yield and stereoselectivity. To the best of our
knowledge, this represents the first use of aliphatic N-Boc-
aldimines for a highly enantioselective direct Mannich reaction
with glycine imines.
Our mechanistic rationale for this reaction is shown in Figure
2. Deprotonation of glycine imine 3b and H-bond engagement of
We further explored the substrate scope of the aliphatic N-
Boc-aldimine Mannich reaction. A major concern with aliphatic
N-Boc-aldimines is the potential for base-promoted enamine
formation, leading to undesired side products. We were able to
avoid competitive N-Boc-aldimine decomposition by the
judicious control of several parameters. Thus, methyl glycinate
3a was selected as the enolate for this screen in order to decrease
the reaction time and allow for lower catalyst loadings.
Additionally, some reactions were carried out with excess N-
Boc-aldimine and at lower temperatures. In this way, an array of
aliphatic N-Boc-aldimines with a variety of substitution patterns
and functional groups were found to be suitable substrates for
this reaction (Table 3).
a
Table 3. Aliphatic Substrate Scope of the Mannich Reaction
Figure 2. Proposed catalytic cycle for the cyclopropenimine-catalyzed
Mannich reaction. The red double-headed arrow indicates conforma-
tional locking of the stereocenter due to steric conflict with a cyclohexyl
ring.
N-Boc imine 2 by the catalyst 1 leads to pre-transition-state
complex 8a or 8b, with subsequent C−C bond formation as the
rate- and enantiodetermining step. This latter hypothesis is
supported by (1) the low reaction rate of the less electrophilic p-
methoxybenzaldimine (Table 2, entry 3) and (2) the significantly
lower rate and diminished enantioselectivity of p-trifluorome-
thylbenzaldimine (entry 5), which can be rationalized as a
consequence of the decreased H-bond acceptor capacity of this
substrate. The precise organization of the transition state,
including the question of whether the enolate is OH-bound (8a)
or NH-bound (8b), is currently under investigation.
Finally, we investigated the performance of catalyst 1 for the
production of preparative amounts of chiral diamino acid
derivatives (Scheme 1 and eq 3). When a 1 mol % loading of
catalyst 1 was used, 26.7 g of diamino ester 9 was prepared (73%
yield, 96:4 dr, 93% ee) in 8 h following imine hydrolysis with
citric acid. Adduct 9 was subsequently converted to several useful
derivatives, including diketopiperazine 10, β-lactam 11, amidine
12, and tripeptide 13, in preparative quantities in three or fewer
operations (Scheme 1). The efficient performance of catalyst 1
a
Yields based on purified products. Diastereomeric ratios (dr) and
b
enantiomeric excesses (ee) were determined by HPLC. 1 gram scale
of glycine imine. 4 equiv of N-Boc-Imine was used. 0 °C. N-Boc-
imine was made in situ with Cs2CO3; 20 mol % catalyst was used. −25
°C.
c
d
e
f
C
dx.doi.org/10.1021/ja407277a | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
(5) (S)-2-(2,3-Bis(dicyclohexylamino)cyclopropenimine)-3-phenyl-
propan-1-ol hydrochloride (1·HCl) can be purchased from Sigma-
Aldrich (product no. L511188).
Scheme 1. Preparative-Scale Synthesis and Derivatization of
Diamine 8
a
(6) For selected examples of enantioselective Mannich reactions using
other donors, see: (a) Zhang, W.-Q.; Cheng, L.-F.; Gong, L.-Z. Angew.
Chem., Int. Ed. 2012, 51, 4085. (b) Liu, X.; Deng, L.; Jiang, X.; Yan, W.;
Liu, C.; Wang, R. Org. Lett. 2010, 12, 876. (c) Uragachi, D.; Ueki, Y.;
Ooi, T. J. Am. Chem. Soc. 2008, 130, 14088. (d) Shi, S.-H.; Huang, F.-P.;
Zhu, P.; Dong, Z.-W.; Hui, X.-P. Org. Lett. 2012, 14, 2010. (e) Cutting,
G. A.; Stainforth, N. E.; John, M. P.; Kociok-Kohn, G.; Willis, M. C. J.
̈
Am. Chem. Soc. 2007, 129, 10632. (f) Li, L.; Ganesh, M.; Seidel, D. J. Am.
Chem. Soc. 2009, 131, 11648. (g) Jiang, J.; Xu, H.-D.; Xi, J.-B.; Ren, B.-Y.;
Lv, F.-P.; Guo, X.; Jiang, L.-Q.; Zhang, Z.-Y.; Hu, W.-H. J. Am. Chem. Soc.
2011, 133, 8428. (h) Chen, X.; Dong, S.; Qiao, Z.; Zhu, Y.; Xie, M.; Lin,
L.; Liu, X.; Feng, X. Chem.Eur. J. 2011, 17, 2583. (i) Lu, G.; Yoshino,
T.; Morimoto, H.; Matsunaga, S.; Shibasaki, M. Angew. Chem., Int. Ed.
2011, 50, 4382. (j) Nakamura, S.; Maeno, Y.; Ohara, M.; Yamamura, A.;
Funahashi, Y.; Shibata, N. Org. Lett. 2012, 14, 2960. (k) Melhado, A. D.;
Amarante, G. W.; Wang, Z. J.; Luparia, M.; Toste, F. D. J. Am. Chem. Soc.
2011, 133, 3517. (l) Chowdari, N. S.; Ahmad, M.; Albertshofer, K.;
Tanaka, F.; Barbas, C. F., III. Org. Lett. 2006, 8, 2839.
a
Conditions: (a) TFA, CH2Cl2, rt; (b) dimethyloxylate, MeOH,
reflux; (c) CBzCl, Na2CO3(aq), PhMe, rt; (d) TMSCl, CH2Cl2, 0 °C,
then tBuMgCl; (e) 4-bromobenzaldehyde, TEA, CH2Cl2, rt, then
NBS, 0 °C; (f) Z-Ala-OH, EDC, HOBt, TEA, CH2Cl2, 0 °C to rt; (g)
Boc-Phe-OH, EDC, HOBt, TEA, CH2Cl2, 0 °C to rt. See the
Supporting Information for full experimental details.
(7) Bernardi, L.; Gothelf, A. S.; Hazell, R. G.; Jørgensen, K. A. J. Org.
Chem. 2003, 68, 2583.
(8) For metal-catalyzed Mannich reactions of glycine imines, see:
(a) Hernan
Eur. J. 2010, 16, 1153. (b) Hernan
́
dez-Toribio, J.; Gom
́
ez Arrayas
́
, R.; Carretero, J. C. Chem.
́
dez-Toribio, J.; Gom
́
ez Arrayas, R.;
́
Carretero, J. C. J. Am. Chem. Soc. 2008, 130, 16150. (c) Liang, G.; Tong,
M. C.; Tao, H.; Wang, C. J. Adv. Synth. Catal. 2010, 352, 1851.
(d) Shang, D.; Liu, Y.; Zhou, X.; Liu, X.; Feng, X. Chem.Eur. J. 2009,
15, 3678. (e) Yan, X. X.; Peng, Q.; Li, Q.; Zhang, K.; Yao, J.; Hou, X. L.;
Wu, Y. D. J. Am. Chem. Soc. 2008, 130, 14362. (f) Hernando, E.; Gom
́
ez
Arrayas, R.; Carretero, J. C. Chem. Commun. 2012, 48, 9622. (g) Salter,
́
M. M.; Kobayashi, J.; Shimizu, Y.; Kobayashi, S. Org. Lett. 2006, 8, 3533.
(9) For recent reviews, see: (a) Ooi, T.; Maruoka, K. Angew. Chem., Int.
Ed. 2007, 46, 4222. (b) Hashimoto, T.; Maruoka, K. Chem. Rev. 2007,
107, 5656. (c) O’Donnell, M. J. Acc. Chem. Res. 2004, 37, 506.
(10) (a) Ooi, T.; Kameda, M.; Fujii, J.; Maruoka, K. Org. Lett. 2004, 6,
2397. (b) Okada, A.; Shibuguchi, T.; Ohshima, T.; Masu, H.;
Yamaguchi, K.; Shibasaki, M. Angew. Chem., Int. Ed. 2005, 44, 4564.
(c) Shibuguchi, T.; Mihara, H.; Kuramochi, A.; Ohshima, T.; Shibasaki,
M. Chem.Asian J. 2007, 2, 794.
coupled with its ready availability should make this chemistry a
practical tool for the preparation of 1,2-diamino stereocenters.
ASSOCIATED CONTENT
* Supporting Information
Experimental procedures and product characterization data. This
material is available free of charge via the Internet at http://pubs.
acs.org.
■
S
(11) (a) Superbases for Organic Synthesis: Guanidines, Amidines,
Phosphazenes and Related Organocatalysts; Ishikawa, T., Ed.; Wiley:
́
Chichester, U.K., 2009. (b) Palomo, C.; Oiarbide, M.; Lopez, R. Chem.
Soc. Rev. 2009, 38, 632. (c) Marcelli, T.; Hiemstra, H. Synthesis 2010,
1229. (d) Leow, D.; Tan, C.-H. Synlett 2010, 1589. (e) Fu, X.; Tan, C.-
H. Chem. Commun. 2011, 47, 8210.
AUTHOR INFORMATION
Corresponding Author
TL2240@Columbia.edu
■
(12) For a review, see: Connon, S. J. Chem. Commun. 2008, 2499.
(13) Zhang, H.; Syed, S.; Barbas, C. F., III. Org. Lett. 2010, 12, 708.
Compound 4 was generated as the opposite enantiomer as that shown in
eq 1.
Notes
The authors declare no competing financial interest.
(14) Notably, a chiral guanidine has been used for enantioselective
Mannich reactions with significantly more acidic fluorenone-derived
glycinate Schiff bases. See: Kobayashi, S.; Yazaki, R.; Seki, K.; Yamashita,
Y. Angew. Chem., Int. Ed. 2008, 47, 5613.
(15) Resubjecting product 4b to the reaction conditions did not lead to
any racemization or change in the diastereomeric ratio.
(16) Using dimethyl malonate as the enolate in place of 3 resulted in a
94% yield of the Mannich product in 2 h, although no enantioselectivity
was observed.
ACKNOWLEDGMENTS
■
Financial support was provided by NIH NIGMS (R01
GM102611). T.H.L. is grateful for an Ely Lilly Grantee Award.
J.S.B. is grateful for NDSEG and NSF graduate fellowships. We
thank the Leighton group (Columbia University) for use of their
instrumentation.
REFERENCES
■
(17) At −78 °C, the reaction stalled at 15% conversion and yielded a
product with 83% ee and 99:1 dr.
(1) (a) Viso, A.; de la Pradilla, R. F.; García, A.; Flores, A. Chem. Rev.
2005, 105, 3167. (b) Viso, A.; de la Pradilla, R. F.; Tortosa, M.; García,
A.; Flores, A. Chem. Rev. 2011, 111, PR1. (c) Lucet, D.; Le Galle, T.;
Mioskowski, C. Angew. Chem., Int. Ed. 1998, 37, 2580.
(2) For a review, see: Gom
Rev. 2009, 38, 1940.
́ ́
ez Arrayas, R.; Carretero, J. C. Chem. Soc.
(3) Bandar, J. S.; Lambert, T. H. J. Am. Chem. Soc. 2012, 134, 5552.
(4) See the Supporting Information for an improved synthesis of 1
directly from pentachlorocyclopropane.
D
dx.doi.org/10.1021/ja407277a | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX