Asymmetric reduction of imines with trichlorosilane, catalyzed by sigamide, an amino acid-derived formamide: Scope and limitations

Article abstract of DOI:10.1021/jo900561h

(Chemical Equation Presented) Enantioselective reduction of ketimines 6-10 with trichlorosilane can be catalyzed by the N-methyl valine-derived Lewis-basic formamide (S)-23 (Sigamide) with high enantioselectivity (≤97% ee) and low catalyst loading (1-5 mol %) at room temperature in toluene. The reaction is efficient with ketimines derived from aromatic amines (aniline and anisidine) and aromatic, heteroaromatic, conjugated, and even nonaromatic ketones 1-5, in which the steric difference between the alkyl groups R¹ and R ² is sufficient. Simple nitrogen heteroaromatics (8a,b,d) exhibit low enantioselectivities due to the competing coordination of the reagent but increased steric hindrance in the vicinity of the nitrogen (8c,e) results in a considerable improvement. Cyclic imines 32d-d exhibited low to modest enantioselectivities.

Full text of DOI:10.1021/jo900561h

pubs.acs.org/joc
Asymmetric Reduction of Imines with Trichlorosilane, Catalyzed by
Sigamide, an Amino Acid-Derived Formamide: Scope and Limitations†
,‡
Andrei V. Malkov,* Kvetoslava Vrankova, Sigitas Stoncius, and Pavel Kocovsky*
ꢀ
ꢁ
ꢁ
ꢀ
Department of Chemistry, WestChem, Joseph Black Building, University of Glasgow, Glasgow G12 8QQ,
Scotland, United Kingdom. ‡Present address: Department of Chemistry, Loughborough University,
Loughborough LE11 3TU, UK.
a.malkov@lboro.ac.uk; pavelk@chem.gla.ac.uk
Received February 25, 2009
Enantioselective reduction of ketimines 6-10 with trichlorosilane can be catalyzed by the N-methyl
valine-derived Lewis-basic formamide (S)-23 (Sigamide) with high enantioselectivity (e97% ee) and
low catalyst loading (1-5 mol %) at room temperature in toluene. The reaction is efficient with
ketimines derived from aromatic amines (aniline and anisidine) and aromatic, heteroaromatic,
conjugated, and even nonaromatic ketones 1-5, in which the steric difference between the alkyl
groups R1 and R2 is sufficient. Simple nitrogen heteroaromatics (8a,b,d) exhibit low enantioselec-
tivities due to the competing coordination of the reagent but increased steric hindrance in the vicinity
of the nitrogen (8c,e) results in a considerable improvement. Cyclic imines 32d-d exhibited low to
modest enantioselectivities.
Introduction
such as hydrogenation,3,6 that are generally tainted by metal
leaching, high pressure, and cost implications. In view of its
simplicity and reliability, this approach may even become an
Enantioselective reduction of imines with trichlorosilane,
catalyzed by chiral, metal-free Lewis bases,1 is gradually
gaining ground as a reliable protocol for obtaining chiral
amines of high enantiomeric purity (Scheme 1). This new
organocatalytic methodology has the promise to successfully
compete with the traditional metal-catalyzed reductions,2-6
(3) For recent reports on catalytic hydrogenation (with Ti, Ir, Rh, and
Ru), see refs 2b-d and the following: (a) Xiao, D.; Zhang, X. Angew. Chem.,
Int. Ed. 2001, 40, 3425. (b) Jiang, X. B.; Minnaard, A. J.; Hessen, B.; Feringa,
B. L.; Duchateau, A. L. L.; Andrien, J. G. O.; Boogers, J. A. F.; de Vries, J. G. Org.
Lett. 2003, 5, 1503. (c) Cobley, C. J.; Henschke, J. P. Adv. Synth. Catal. 2003,
345, 195. (d) Okuda, J.; Verch, S.; St€urmer, R.; Spaniol, T. S. J. Organomet.
† Dedicated to Prof. Josef Michl on the occasion of his 70th birthday.
Chem. 2000, 605, 55. (e) Guiu, E.; Munoz, B.; Castillon, S.; Claver, C. Adv.
Synth. Catal. 2003, 345, 169. (f) Cobbley, C. J.; Foucher, E.; Lecouve, J.-P.;
Lennon, I. C.; Ramsden, J. A.; Thominot, G. Tetrahedron: Asymmetry 2003, 14,
3431. (g) Chi, Y.; Zhou, Y. G.; Zhang, X. J. Org. Chem. 2003, 68, 4120.
~
ꢀ
ꢁ
(1) (a) Kocovskꢀy, P.; Malkov, A. V. Chiral Lewis Bases as Catalysts. In
Enantioselective Organocatalysis; Dalko, P. I., Ed.; Wiley-VCH: Weinheim,
Germany, 2007; p 255. (b) Kagan, H. Organocatalytic Enantioselective Reduction
of Olefins, Ketones, and Imines. In Enantioselective Organocatalysis; Dalko, P.
I., Ed.; Wiley-VCH: Weinheim, Germany, 2007; p 391.
(2) For a general overview of the reduction of imines, see the following:
(a) Morrison, J. D. Asymmetric Synthesis; Academic: New York, 1983; Vol. 2.
(b) Noyori, R. Asymmetric Catalysis in Organic Synthesis; Wiley & Sons: New
York, 1994. (c) Ojima, I. Catalytic Asymmetric Synthesis, 2nd ed.; J. Wiley and
Sons: New York, 2000. (d) James, B. R. Catal. Today 1997, 37, 209.
(e) Kobayashi, S.; Ishitani, H. Chem. Rev. 1999, 99, 1069. (f) Cho, B. T.
Tetrahedron 2006, 62, 7621. (g) Jacobsen, E. N.; Pfaltz, A.; Yamamoto, H.
Comprehensive Asymmetric Catalysis; Springer: New York, 1999; Vols. I-III.
ꢀ
(h) Bozeio, A. A.; Pytkowicz, J.; Cte, A.; Charette, A. B. J. Am. Chem. Soc. 2003,
125, 14260. (i) Trifonova, A.; Diesen, J. S.; Chapman, C. J.; Andersson, P. G. Org.
Lett. 2004, 6, 3825. (j) Zhu, S.-F.; Xie, J.-B.; Zhang, Y.-Z.; Li, S.; Zhou, Q.-L. J.
Am. Chem. Soc. 2006, 128, 12886. (k) Trifonova, A.; Diesen, J. S.; Andersson, P.
G. Chem.;Eur. J. 2006, 12, 2318. (l) Cheruku, P.; Church, T. L.; Andersson, P.
G. Chem. Asian J. 2008, 3, 1390. For Rh-catalyzed hydrogenation of enamides,
see the following: (m) Hu, X.-P.; Zheng, Z. Org. Lett. 2004, 6, 3585. (n) Yiang,
Q.; Gao, W.; Deng, J.; Zhang, X. Angew. Chem., Int. Ed. 2006, 45, 2832. (o) Li,
C.; Wang, C.; Villa-Macros, B.; Xiao, J. J. Am. Chem. Soc. 2008, 130, 14450.
(p) Moessne, C.; Bolm, C. Angew. Chem., Int. Ed. 2005, 44, 7564.
DOI: 10.1021/jo900561h
r
Published on Web 06/16/2009
J. Org. Chem. 2009, 74, 5839–5849 5839
2009 American Chemical Society

Malkov et al.
JOCArticle
SCHEME 1. Preparation of Selected Ketimines and Their
In the past few years, we have developed a series of Lewis-
basic organocatalysts for the reduction of imines with
Cl3SiH.9,10 Those, based on the scaffold derived from
N-methyl valine (Chart 1) proved to be most promising. A
detailed investigation of the structure-activity relationship
of the catalysts, using a set of model imines 6, has led to the
following conclusions:9b (1) Valine’s isopropyl group, as in
16, represents an optimum in terms of enantioselectivity
(e95% ee). The cyclohexyl analogue, derived from cyclo-
hexyl glycine, exhibited almost identical selectivities,
whereas other catalysts with t-Bu, PhCH2, Ph, and Me in
place of the original i-Pr proved less enantioselective.9b,11 (2)
The N-methyl group is crucial as the corresponding NH
derivative exhibited low enantioselectivity (e35% ee).9b (3)
The formamide function at the N-terminus of valine is
another key factor; the corresponding acetamide, trifluoro-
acetamide, carbamate, and urea derivatives reacted slug-
gishly and gave racemic products.9b (4) The carboxyl termi-
nus of the parent valine needs to be converted into an amide
with a primary aromatic amine.9 Amides derived from
secondary aromatic amines (e.g., MeNHPh), as well as their
nonaromatic congeners (e.g., that derived from BuNH2),
proved inferior.9b (5) Anilide 16, our first catalyst, exhibited
good enantioselectivities (e90% ee); introduction of alkyl
groups into the anilide moiety, as in 17 (Kenamide) and 18,
had a small but noticeable effect (e92% ee).9b,c,h Sigamide
(23), with two tert-butyl groups, was identified as an opti-
mum (e95% ee at rt with 1-5 mol % loading).9c Other
substituents, such as 3,5-dimethoxy (19) and 3,5-
di(trifluoromethyl) (20), had a minor negative effect on the
enantioselectivity.9b By appending a fluorous tag (21)9c or by
anchoring the catalyst to a resin (22),9e nanoparticle,9f
soluble polymer,9i or a dendron,9j we have simplified the
isolation of the product with little effect on the reaction
efficiency.
Asymmetric Reductiona
aFor R1, R2, and Ar, see Table 1.
attractive alternative to the existing enzymatic methods for
amine production,7 and can complement another organoca-
talytic protocol, based on the biomimetic reduction with
Hantzsch ester.8
(4) For hydrosilylation, see, e.g.: (a) Reding, M. T.; Buchwald, S. L.
J. Org. Chem. 1998, 63, 6344. (b) Verdaguer, X.; Lange, U. E. W.; Buchwald,
S. L. Angew. Chem., Int. Ed. 1998, 37, 1103. (c) Hansen, M. C.; Buchwald, S.
L. Org. Lett. 2000, 2, 713. (d) Vedejs, E.; Trapencieris, P.; Suna, E. J. Org.
Chem. 1999, 64, 6724. (e) Nishikori, H.; Yoshihara, R.; Hosomi, A. Synlett
2003, 561. (f) Lipshutz, B. H.; Noson, K.; Chrisman, W. J. Am. Chem. Soc. 2001,
123, 12917. (g) Lipshutz, B. H.; Shimizu, H. Angew. Chem., Int. Ed. 2004, 43,
2228. (h) Lipshutz, B. H.; Frieman, B. A.; Tomaso, A. E. Angew. Chem., Int. Ed.
2006, 45, 1259.
€
(5) For transfer hydrogenation, see, e.g.: (a) Samec, J. S. M.; Backvall,
J.-E. Chem.;Eur. J. 2002, 8, 2955. (b) Kadyrov, R.; Riermeier, T. H. Angew.
€
Chem., Int. Ed. 2003, 42, 5472. (c) Aberg, J. B.; Samec, J. S. M.; Backvall,
J.-E. Chem. Commun. 2006, 2771. For an overview, see: (d) Samec, J. S. M.;
€
Backvall, J.-E.; Andersson, P. G.; Brandt, P. Chem. Soc. Rev. 2006, 35, 237.
For the kinetic resolution using transfer hydrogenation combined with an
€
ezymatic reaction, see: (e) Paetzold, J.; Backvall, J.-E. J. Am. Chem. Soc.
2005, 127, 17620. (f) Samec, J. S. M.; Ell, A. H.; Aberg, J. B.; Privalov, T.;
ꢀ
€
Eriksson, L.; Backvall, J.-E. J. Am. Chem. Soc. 2006, 128, 14293.
(g) Privalov, T.; Samec, J. S. M.; Backvall, J.-E. Organometallics 2007, 26,
€
€
2840. (h) Hoben, C. E.; Kanupp, L.; Backvall, J.-E. Tetrahedron Lett. 2008,
49, 977. (i) Matharu, D. S.; Martins, J. E. D.; Wills, M. Chem.;Asian J.
2008, 3, 1374. For reductive amination (HCO2H, Ru), see: (j) Williams,
G. D.; Pike, R. A.; Wade, C. E.; Wills, M. Org. Lett. 2003, 5, 4227.
(6) Schinder, P.; Kock, G.; Pretot, R.; Wang, G.; Bohnen, F. M.; Kruger,
C.; Pfaltz, A. Chem.;Eur. J. 1997, 3, 887.
(7) For enzymatic approaches, including reduction, see, e.g.: (a) Alexeeva,
M.; Enright, A.; Dawson, M. J.; Mahmoidian, M.; Turner, N. J. Angew.
Chem., Int. Ed. 2002, 41, 3177. (b) Carr, R.; Alexeeva, M.; Dawson, M. J.;
Several other groups have participated in the development
of this methodology (Chart 2).12-14 Thus, Matsumura was
actually the first to demonstrate a promising enantioselec-
tivity in the reduction of 6a with the proline-derived for-
mamide 24 as catalyst (e68% ee).12a More recently, Sun13
showed that expansion of the five-membered ring of the
proline moiety as in the six-membered pipecolinic derivative
25, or in its piperazidine analogue 26, and further elabora-
tion of the carboxylic terminus of the parent amino acid, had
a beneficial effect on the enantioselectivity, which could now
be increased up to 95% ee (at 0 °C with 10 mol % catalyst
loading).10d,13b,c On the other hand, the proline-derived
ꢀ
Gotor-Fernandez, V.; Huphrey, C. E.; Turner, N. J. ChemBioChem 2005, 6,
637. (c) Dunsmore, C. J.; Carr, R.; Fleming, T.; Turner, N. J. J. Am. Chem.
Soc. 2006, 128, 2224. For a brief overview, see: (d) Alexeeva, M.; Carr, R.;
Turner, N. J. Org. Biomol. Chem. 2003, 1, 4133.
(8) (a) Singh, S.; Batra, U. K. Indian J. Chem., Sect. B 1989, 28, 1.
(b) Rueping, M.; Sugiono, E.; Azap, C.; Theissmann, T.; Bolte, M. Org. Lett.
2005, 7, 3781. (c) Rueping, M.; Antonchick, A. P.; Theissmann, T. Angew.
Chem., Int. Ed. 2006, 45, 3683. (d) Rueping, M.; Antonchik, A. P.; Theissmann,
T. Angew. Chem., Int. Ed. 2006, 45, 6751. (e) Hoffman, S.; Seayad, A. M.; List,
B. Angew. Chem., Int. Ed. 2005, 44, 7424. (f) Hoffmann, S.; Nicoletti, M.; List,
B. J. Am. Chem. Soc. 2006, 128, 13074. (g) Zhou, J.; List, B. J. Am. Chem. Soc.
2007, 129, 7498. (h) Storer, R. I.; Carrera, D. E.; Ni, Y.; MacMillan, D. W. C.
J. Am. Chem. Soc. 2006, 128, 84. (i) Ouellet, S. G.; Walji, A. M.; MacMillan, D.
W. C. Acc. Chem. Res. 2007, 40, 1327.
ꢁ
(9) (a) Malkov, A. V.; Mariani, A.; MacDougall, K. N.; Kocovsky, P.
Org. Lett. 2004, 6, 2253. (b) Malkov, A. V.; Stoncius, S.; MacDougall, K. N.;
ꢀ
ꢁ
ꢁ
Mariani, A.; McGeoch, G. D.; Kocovsky, P. Tetrahedron 2006, 62, 264.
(c) Malkov, A. V.; Figlus, M.; Stoncius, S.; Kocovsky, P. J. Org. Chem. 2007,
ꢀ
ꢁ
ꢁ
ꢀ
(11) For catalysts based on the proline scaffold, see below (refs 12-14).
(12) (a) Iwasaki, F.; Onomura, O.; Mishima, K.; Kanematsu, T.; Maki,
T.; Matsumura, Y. Tetrahedron Lett. 2001, 42, 2525. (b) Onomura, O.;
Kouchi, Y.; Iwasaki, F.; Matsumura, Y. Tetrahedron Lett. 2006, 47, 3751.
(13) (a) Wang, Z.; Ye, X.; Wei, S.; Wu, P.; Zhang, A.; Sun, J. Org. Lett.
2006, 8, 999. (b) Wang, Z.; Cheng, M.; Wu, P.; Wei, S.; Sun, J. Org. Lett.
2006, 8, 3045. (c) Wu, P.; Wang, Z.; Cheng, M.; Zhou, L.; Sun, J. Tetrahedron
2008, 64, 11304. (d) Wang, Z.; Wei, S.; Wang, C.; Sun, J. Tetrahedron:
Asymmetry 2007, 18, 705. (e) Pei, D.; Wang, Z.; Wei, S.; Zhang, Y.; Sun, J.
Org. Lett. 2006, 8, 5913. (f) Wang, C.; Wu, X.; Zhou, L.; Sun, J. Chem.;Eur.
J. 2008, 14, 8789. (g) Pei, D.; Zhang, Y.; Wei, S.; Wang, M.; Sun, J. Adv.
Synth. Catal. 2008, 350, 619.
(14) (a) Zheng, H.; Deng, J.; Lin, W.; Zhang, X. Tetrahedron Lett. 2007,
48, 7934. (b) Zheng, H.-J.; Chen, W. B.; Wu, Z.-J.; Deng, J.-G.; Lin, W.-Q.;
Yuan, W.-C.; Zhang, X.-M. Chem.;Eur. J. 2008, 14, 9864. See also:
(c) Guizzetti, S.; Benaglia, M.; Cozzi, F.; Rossi, S.; Celentano, G. Chirality
2009, 21, 233. (d) Gautier, F. M.; Jones, S.; Martin, S. J. Org. Biomol. Chem.
2009, 7, 229.
ꢁ
72, 1315. (d) Malkov, A. V.; Stoncius, S.; Kocovsky, P. Angew. Chem., Int.
Ed. 2007, 46, 3722. (e) Malkov, A. V.; Figlus, M.; Kocovsky, P. J. Org. Chem.
ꢁ
ꢀ
ꢁ
ꢀ
2008, 73, 3985. (f) Malkov, A. V.; Figlus, M.; Cooke, G.; Caldwell, S. T.; Rabani,
ꢁ
G.; Prestly, M. R.; Kocovskꢀy, P. Org. Biomol. Chem. 2009, 1878. (g) Malkov, A.
ꢁ
ꢀ
ꢁ
V.; Stoncius, S.; Vrankova, K.; Arndt, M.; Kocovskꢀy, P. Chem.;Eur. J. 2008, 14,
ꢀ
ꢁ
ꢁ
8082. (h) Malkov, A. V.; Vrankova, K.; Sigerson, R.; Stoncius, S.; Kocovskꢀy, P.
manuscript in preparation. (i) Malkov, A. V.; Figlus, M.; Prestly, M. R.; Rabani,
ꢁ
ꢀ
G.; Cooke, G.; Kocovsky, P. manuscript in preparation. (j) Cooke, G.; Figlus, M.;
ꢁ
Caldwell, S. T.; Walas, D.; Sanyal, A.; Yesilbag, G.; Malkov, A. V.; Kocovskꢀy, P.
manuscript in preparation.
(10) For an analogous reduction of ketones with Cl3SiH, catalyzed by
metal-free Lewis bases, see : (a) Iwasaki, F.; Onomura, O.; Mishima, K.;
Maki, T.; Matsumura, Y. Tetrahedron Lett. 1999, 40, 7507. (b) Matsumura,
Y.; Ogura, K.; Kouchi, Y.; Iwasaki, F.; Onomura, O. Org. Lett. 2006, 8,
ꢁ
3789. (c) Malkov, A. V.; Stewart Liddon, A. J. P.; Ramırez-Lopez, P.;
Bendova, L.; Haigh, D.; Kocovsky, P. Angew. Chem., Int. Ed. 2006, 45,
ꢀ
ꢁ
ꢀ
1432. (d) Zhou, L.; Wang, L.; Wei, L.; Sun, J. Chem. Commun. 2007, 2977.
5840 J. Org. Chem. Vol. 74, No. 16, 2009

Malkov et al.
JOCArticle
CHART 1. Catalysts for the Asymmetric Reduction of Imines
represents the most recent addition to this family; here, he
replaced the Lewis-basic formyl group of the proline-derived
catalyst 24 by the t-BuSO moiety, which resulted in the
enhanced enantioselectivities (e97% ee at 0 °C with
10 mol % catalyst loading).13f An analogous bissulfinamide,
which apparently coordinates to trichlorosilane in a biden-
tate fashion, exhibited comparable enantioselectivities
(e96% ee at -20 °C with 10 mol % catalyst loading).13g
Herein, we present an orchestration of our earlier work,9
using Sigamide (23), our champion catalyst, which is now
commercially available. The substrate portfolio, originally
limited mainly to the imines derived from acetophenone and
its congeners,9,12-14 is substantially expanded to the hetero-
cyclic and aliphatic realm; tolerance of a variety of functional
groups and substitution patterns is also demonstrated.
CHART 2. Other Catalysts for the Asymmetric Reduction of
Imines
Results and Discussion
In our previous studies,9 we have identified the N-methyl
valine-derived formamide 23 (Sigamide) as the organocata-
lyst of choice for the reduction of aromatic ketimines 6a-e
(Scheme 1 and Table 1). High levels of enantioselectivities
were attained (91-94% ee), typically with 5 mol % catalyst
loading (Table 1, entries 3-8). In fact, the loading can be
reduced to 1 mol % without any loss of enantioselectivity
(entry 5)9c but further lowering turned out to lead to unreli-
able results, as the uncatalyzed background reaction
(vide infra) began to compete. Sigamide 23 exhibited
slightly higher enantioselectivities than its dimethyl analogue
Kenamide 17 (Table 1, compare entries 1 and 2 with 3);
therefore, Sigamide was used as the sole catalyst in the
present study. Toluene was identified previously by us as
an optimal solvent, with slightly higher enantioselectivities
than those attained with the less environmentally friendly
CHCl3 or CH2Cl2.9a-d,16 Therefore, we felt no need to
explore other solvents again. Sufficiently high enantioselec-
tivities have been found for the reductions run at
room temperature, with only a marginal improvement
at 0 or -20 °C or lower (at the expense of the reaction rate);9
hence, ambient temperature (15-20 °C) was maintained for
all reductions throughout this study.
Synthesis of Ketimines 6-10. The preparation of ketimines
6-8 (Scheme 1) from the corresponding methylketones 1-3
and aniline or anisidine, using Dean-Stark conditions in the
presence of p-TsOH as catalyst or using molecular sieves to
remove water, was mostly uneventful and afforded the
required imines in good isolated yields (no optimization
was attempted): 6a (68%),3k,6 6b (66%),17 6c (53%),9c 6d
(68%),9b,c,10b 6e (60%),9b,c,10b 6f (87%), 6g (64%),4c,9b 6h
(71%),9e 7a (55%), 7b (22%), 7c (31%), 7d (38%),9c 8a
(95%), 8b (96%), 8c (∼57%), 8d (95%), 8e (73%), 8f
(56%), 8g (41%), 8h (78%), 8i (74%), 8j (74%), 8k (72%),
8l (74%), 8m (31%), 8n (57%), and 8o (40%). Other sterically
undemanding arylalkyl ketones were also readily converted
into the corresponding imines 9a (48%), 9b (16%), 9c (40%),
9d (38%), 9e (72%), 9f (27%), 9g (58%), 34c (69%),3p
and 34d (70%)4c under the same conditions. However, the
dimer analogous to 24, expected to offer an extensive chela-
tion of the reagent, did not provide any particular advantage
(e77% ee).13d,15 An interesting catalyst with a novel design,
featuring a sulfinamide group as the chiral element (27), was
also successful (e92% ee at -20 °C with 20 mol % catalyst
loading).13e On the other hand, replacement of the forma-
mide moiety in the original proline scaffold by the R-picolinic
amide, in conjunction with the introduction of a hindered
tertiary hydroxyl in place of the amide function of proline
(28), did not lead to any improvement (e75% ee).12b Never-
theless, the latter design presented one interesting structural
feature to be further elaborated, namely the R-picolinic
amide moiety and the hydroxy group. Being apparently
inspired by this new turn, Zhang has recently developed
the R-picolinic amide 29, in which he employed ephedrine as
the chiral scaffold.14 The latter catalyst exhibited high en-
antioselectivities (e93% at -10 °C with 20 mol % catalyst
loading) and seems to have the promise to work even with
those substrates with which other catalysts were either
inefficient or failed. Sulfinamide 30, introduced by Sun,
(16) Other solvents, such as MeCN, THF, and ether, proved inferior,9a-d
while the Lewis-basic DMF, DMSO, or HMPA could not be used as these
would interfere with the catalyst coordination of the silicon. Protic solvents
would decompose the reagent, so that they were also excluded.
(17) Barluenga, A.; Aznar, F. Synthesis 1975, 704.
(15) The heptacoordinate silicon, portrayed in the proposed transi-
tion state by Sun,13d seems unlikely.
J. Org. Chem. Vol. 74, No. 16, 2009 5841

Malkov et al.
JOCArticle
TABLE 1. Reduction of Ketimines 6-10 with Trichlorosilane to Give Amines 11-15, Catalyzed by the N-Methyl Valine-Derived Formamides (S)-17 and (S)-23a
entry catalyst (mol %) imine 6-10 Ar
R1 R2 amine 11-15 yield (%)b 11-15c % eed
1
2
3
4
5
6
7
8
17 (10)
17 (10)
23 (5)
23 (2.5)
23 (1)
23 (5)
23 (5)
23 (5)
23 (5)
17 (10)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
17 (10)
23 (5)
23 (5)
23 (5)
23 (5)
23 (5)
6a
6b
6b
6b
6b
6c
6d
6e
6f
6g
6h
7a
7b
7c
7d
7e
8a
8b
8c
8d
8e
8f
8g
8h
8i
8j
8k
8l
8m
8n
8o
9a
9b
9c
9d
9e
9f
9g
9h
9i
9j
9k
9l
9m
9n
10a
10b
10c
10d
10e
10f
10g
10h
10i
10j
10k
10l
10m
10n
10o
10p
32a
32b
32c
32d
Ph
Ph
Ph
Ph
Ph
Ph
Me
Me
Me
Me
Me
Me
Me
Me
Ph
11a
81
85
95
92
92
89
92
91
72
90
93
94
62
86
83
92e
91e
94f
94f
93f
92f
92f
91f
93
4-MeO-C6H4 11b
4-MeO-C6H4 11b
4-MeO-C6H4 11b
4-MeO-C6H4 11b
3,5-Me2C6H3 11c
4-MeO-C6H4 11d
4-MeO-C6H4 11e
4-MeO-C6H4 11f
4-CF3C6H4
4-MeOC6H4
3-(t-Bu)Me2SiO-C6H4 Me
2-MeC6H4
2-naphthyl
(E)-PhCHdCH
(E)-PhCHdC(Me)
c-hexyl
i-Pr
t-Bu
2-pyridyl
4-pyridyl
2,6-(i-Pr)2-4-pyridyl
2-thiazolyl
5-(i-Pr)-2-thiazolyl
1-Me-2-indolyl
thiophen-2-yl
2-furyl
5-Me-2-furyl
5-(Me3Si)-2-furyl
5-(EtO2C)-2-furyl
2-benzofuryl
3-furyl
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Et
Ph
11g
92e
92f
81f
84
4-MeO-C6H4 11h
4-MeO-C6H4 12a
4-MeO-C6H4 12b
4-MeO-C6H4 12c
4-MeO-C6H4 12d
4-MeO-C6H4 12e
4-MeO-C6H4 13a
4-MeO-C6H4 13b
4-MeO-C6H4 13c
4-MeO-C6H4 13d
4-MeO-C6H4 13e
4-MeO-C6H4 13f
4-MeO-C6H4 13g
4-MeO-C6H4 13h
4-MeO-C6H4 13i
4-MeO-C6H4 13j
4-MeO-C6H4 13k
4-MeO-C6H4 13l
4-MeO-C6H4 13m
4-MeO-C6H4 13n
4-MeO-C6H4 13o
4-MeO-C6H4 14a
4-MeO-C6H4 14b
4-MeO-C6H4 14c
4-MeO-C6H4 14d
4-MeO-C6H4 14e
4-MeO-C6H4 14f
85f
62f
39 [38g]
7
21
78
63 [92g]
72
85
∼68i
83
71
13
34
91
89
77
77
62 [57h]
86
62
56 [62h]
45
63
62
79
60
84
90
81
64
62
98
99
95
69
98
87
75
70
70
77
65
91
92
84
82
90
95
95
96
97
95
94
76
3-benzofuryl
2,5-Me2-3-furyl
Ph
Ph
(CH2)3CHdCH2
(CH2)3CHdCH2
(CH2)3CHdCH2
(CH2)2CHdCH2
(CH2)2Ph
3-Cl-C6H4
3-MeO-C6H4
3-MeO-C6H4
Ph
3-(t-Bu)Me2SiO-C6H4 (CH2)2(3,4,5-MeO-C6H2) 4-MeO-C6H4 14g
Ph
Ph
Ph
Ph
Ph
2-furyl
4-MeO-C6H4
Ph
4-Cl-C6H4
4-F-C6H4
2-Cl-C6H4
2-Cl-C6H4
2-Cl-C6H4
4-MeO-C6H4
3-MeO-C6H4
2-naphthyl
Ph
i-Pr
c-Pr
c-Bu
c-hexyl
t-Bu
i-Pr
4-MeO-C6H4 14h
4-MeO-C6H4 14i
4-MeO-C6H4 14j
4-MeO-C6H4 14k
4-MeO-C6H4 14l
4-MeO-C6H4 14m
4-MeO-C6H4 14n
4-MeO-C6H4 15a
4-MeO-C6H4 15b
4-MeO-C6H4 15c
4-MeO-C6H4 15d
4-Cl-C6H4
4-F-C6H4
4-MeO-C6H4 15g
4-MeO-C6H4 15h
4-MeO-C6H4 15i
83
18 [46g]
∼75i
79
10 [10g]
85
4-CF3-C6H4
CH2Cl
CH2Cl
CH2Cl
CH2Cl
CH2Cl
CH2Cl
CH2Cl
CH2Cl
CH2Cl
CH2CN
CH2CN
CH2CO2Et
CH2CO2Et
CH2CO2Et
CO2Me
(CH2)2CO2Me
na
6
98j
87j
92j
54j,m
87j,m
86j,m
86j,m
84j,m
92j,m
97n
75n
98n
95n
80n
69h
38r
69
96k,l
91k,l
94k,l
96k,l
95k,l
95k,l
91k,l
92k,l
91k,l
87o,p
87p
89p,q
90p
88p
59e,l
88
15e
15f
Ph
15j
Ph
Ph
4-MeO-C6H4 15k
4-MeO-C6H4 15l
4-MeO-C6H4 15m
4-MeO-C6H4 15n
Ph
4-MeO-C6H4 15p
na
na
na
na
4-F-C6H4
4-MeO-C6H4
Ph
Ph
na
na
na
na
15o
33a
33b
33c
33d
25s
46s
0
na
na
na
91
89
83
10
aThe reaction was carried out at 0.2 mmol scale with 2.0 equiv of Cl3SiH at 18 °C for 16 h in toluene unless stated otherwise. bIsolated yield. cThe
absolute configuration of the amines was established by comparison of their optical rotation (measured in CHCl3) with the literature data (see the
Experimental Section) and/or by comparison of their HPLC behavior with that of authentic samples.9 Amines 11a,b,d,e,h and 15k,l were (S)-configured;
the configuration of the remaining amines is assumed to be (S) in analogy with the rest of the series and our previous experiments. dDetermined by chiral
HPLC. eReferences 9a and 9b. fReference 9c. gIn the presence of AcOH (1 equiv). hThe reaction was carried out at -20 °C for 24 h. iEstimated yield
(NMR) as the starting imine could not be isolated as a pure compound. jThe reaction was carried out for 24 h. kSee ref 9d for preliminary results. lThe
product was (R)-configured; however, note the change in the preference of the substituents in the Cahn-Ingold-Prelog system. mThe imine was
generated in situ and reduced without isolation. nThe reaction was carried out for 48 h in the presence of acetic acid (1 equiv). oA single crystallization
from hexane afforded 15j of 99.9% ee. pReference 9g. qA single crystallization from hexane afforded 15l of 99.8% ee. r2-Methoxy-1-(40-methoxyphenyl)-
5-phenyl-1H-pyrrole was isolated as the major byproduct (27%), arising by a ring-closure of 15p. sFor the absolute configuration, see the text and ref 24.
5842 J. Org. Chem. Vol. 74, No. 16, 2009

Malkov et al.
JOCArticle
sterically more hindered ketones, such as phenylisopropyl
ketone, proved less reactive, so that imine 9h was obtained in
only 29% yield and 9m in ∼10% yield (not isolated as a pure
compound). Therefore, TiCl4 was employed instead of
p-TsOH18 for ketones with c-Pr, c-Bu, c-Hex, t-Bu, and
two aryl groups to obtain the sterically more congested
imines 7e (75%), 9i (52%), 9j (63%), 9k (53%), 9l (68%),
and 9n (69%). The R-chloro imines 10a (68%),9d 10b
(66%),9d and 10c (46%)9d were prepared either by using
the standard p-TsOH/Dean-Stark method or the molecular
sieves. However, their electron-rich or sterically congested
counterparts 10d-i proved rather unstable and difficult to
prepare in a pure state; therefore, they were generated in situ
(with an excess of the ketone19) prior to the reduction and
were not isolated. The remaining imines 10j,9g 10k,9g 10l,9g
10m,9g and 10n9g were obtained in their tautomeric enamine
forms 31j (62%),9g 31k (51%),9g 31l (66%),9g 31m (72%),9g
and 31n (72%)9g (Scheme 2) by the reaction of anisidine or
aniline with the corresponding ketonitriles in acetic acid9g,20
or from ketoesters in ethanol.21 Imines 10o (87%)9b,12a and
10p (60%) were prepared from the corresponding ketones by
using the Dean-Stark method. Interestingly, attempted
synthesis or in situ generation of the imines derived from
R-methoxyacetophenone and R-phthalimidoacetophenone
(and anisidine) failed.
overnight (a compromise to attain both good reaction rates
and enantioselectivities), mostly with Sigamide (23) as cata-
lyst at 5 mol % loading (although 1 mol % was also shown to
be equally effective9c) under an argon atmosphere. In some
cases, acetic acid (e1 equiv) was added to the reaction
mixture (vide infra).
Imines 6a-h, derived from acetophenone and its conge-
ners, afforded the corresponding amines 11a-h in high yields
and >90% ee (entries 1-11). Sigamide (23) was found to be
a slightly more efficient catalyst than Kenamide (17)
(compare entries 2 and 3). Electronic properties of the imine
had no effect on the reduction, as documented by the extreme
examples of the electron-poor and electron-rich imines 6d
and 6e (Table 1, entries 7 and 8). Similarly, an increased steric
hindrance, exercised by an ortho substituent (6g), proved to
be of no consequence (entry 10).
Distance separation between the aromatic nucleus and the
iminemoiety, asinthe conjugatedcinnamylderivatives7aand
7b, had a minor negative effect on the enantioselectivity (81/
84% ee; entries 12 and 13); a noticeable deceleration of the
reaction, reflected in the lower isolated yield, was observed for
the more sterically hindered methyl homologue 7b.
A complete removal of the conjugated π-system, as in the
cyclohexyl analogue 7c, also resulted in a minor decrease in
enantioselectivity (to 85% ee, entry 14) as compared to the
phenyl analogue 6b (94% ee, entry 3). However, it is perti-
nent to note that this decrease was found to be more
substantial when the less bulky catalyst 16 was employed
(37% ee at rt and 59% ee at -20 °C),9b demonstrating the
superiority of Sigamide (23). With the less sterically hindered
isopropyl derivative 7d, the enantioselectivity dropped to
62% ee (entry 15). On the other hand, the sterically much
more congested tert-butyl derivative 7e reacted sluggishly
and the enantioselectivity was found to decrease further
(entry 16).
Introduction of a heteroatom into the aromatic system of
model compounds is always an ultimate goal in view of the
role that heterocyclic compounds play in the pharmaceutical
industry. Therefore, the set of the imines to be investigated
was extended to representative examples of the heteroaro-
matic realm. The pyridyl derivatives 8a,b were reduced
cleanly but the products turned out to be almost racemic
(entries 17 and 18). In another study, we have shown that the
pyridine nitrogen is capable of coordinating the silicon of
Cl3SiH;10c hence, the lack of enantioselection in the case of
8a,b can be attributed to this type of coordination, which is
nonchiral and can compete with the coordination to the
chiral catalyst. The same effect can account for the very low
enantioselectivity observed for the thiazole derivative 8d
(entry 20). To reduce or eliminate the undesired coordina-
tion, a steric bulk was built around the nitrogen of the
substrate, as in the 2,6-diisopropylpyridine derivative 8c.
This strategy proved to have a positive impact, as the
enantioselectivity in the latter case was increased to 78% ee
(entry 19). Introduction of just one isopropyl group, as in 8e,
was less effective (34% ee, entry 21). In contrast to the
pyridine-derived imines, the N-methyl-2-indolyl derivative
8f with a noncoordinating nitrogen atom again exhibited
high enantioselectivity (91% ee; entry 22).
In most cases, anisidine, rather than aniline, was employed
in the imine synthesis. While the reactivity of these two types
of imines hardly differs,9 the p-methoxyphenyl group has the
advantage that it can be oxidatively removed from the
resulting amine to produce a primary amine,9d which renders
the whole protocol more versatile.
The configuration at the imine double bond was predo-
1
minantly (E) (7:1-10:1), as revealed by H NMR spectro-
scopy for 6-8 and 9a-g; for the chloroimines 10a-i the
same predominant configuration needs to be denoted as (Z)
due to the change in substituent preference in the Cahn-
Ingold-Prelog system. Mixtures of (E/Z) isomers were
isolated when the R2 group was bulky, in particular for
9h-m (5:2-5:3), whereas 9n was a ∼3:2 mixture; for details,
see the Experimental Section. Imines 10j-n preferentially
exist in their enamine form 31j-n (Scheme 2). The β-enam-
ino esters 31l-n were obtained as pure (Z) isomers, which are
apparently more stable owing to an intramolecular hydrogen
bonding between the N-H and the ester carbonyl group. By
contrast, the β-enamino nitriles 31j,k readily equilibrate to a
mixture of (Z/E) isomers, e.g., in CDCl3 solutions.9g The
isomerization is facilitated by a trace of Brønsted acids and
apparently occurs via the imine tautomer 10j-n, which is
assumed to be the actual species reduced by trichlorosilane
(vide infra).9g
Reduction of Ketimines 6-10. The reduction of ketimines
(Scheme 1 and Table 1) was carried out under our standard
conditions,9 i.e., with 2 equiv of Cl3SiH (this excess could be
considerably lowered for larger-scale operations) in toluene
(an optimized solvent) at room temperature (15-20 °C)
ꢀ
(18) For the method, see: Sulmon, P.; De Kimpe, N.; Verhe, R.; De
Buyck, L.; Schamp, N. Synthesis 1985, 192.
(19) The excess of the ketone does not interfere with the imine reduction,
as ketones do not undergo reduction with Cl3SiH and this type of catalyst.9d
(20) This procedure was a further modification of the original protocol:
Rao, V. V. R.; Wentrup, C. J. Chem. Soc., Perkin Trans. 1 2002, 1232.
(21) Dai, Q.; Yang, W.; Zhang, X. Org. Lett. 2005, 7, 5343.
Sulfur as a heteroatom is apparently free of negative
effects, as shown by the thiophen-2-yl derivative 8g, where
the enantioselectivity (89% ee, entry 23) was close to that
J. Org. Chem. Vol. 74, No. 16, 2009 5843

Malkov et al.
JOCArticle
observed for the acetophenone-derived imines (entry 3).
Furan- and benzofuran-derived imines 8h-n were found to
be reduced with the efficiency laying between that of
the sulfur and nitrogen heterocycles (45-77% ee, entries
24-30); the 3-furyl derivative 8m exhibited a significantly
higher level of asymmetric induction (77% ee, entry 29) than
its 2-furyl isomer 8h (56% and 62% ee at rt and at -20 °C,
respectively, entry 24). Again, increasing the steric bulk
had a positive effect, as documented by the dimethyl deriva-
tive 8o (91% ee, entry 31).
SCHEME 2. Reaction of 10j-n to 31j-n
2
X = CN, CO Et (see Table 1).
generate the imine in situ, using an excess of the chloro
ketone, to make sure that all the anisidine was consumed in
the reaction, as its presence in the reduction step was
known9d to be detrimental to the enantioselectivity. Details
of this two-step one-pot protocol were revealed in our
preliminary communication.9d
The imines investigated so far were all derived from
acetophenone and its congeners, i.e., from aryl/heteroaryl
methyl ketones. Naturally, it was of interest to vary the
original methyl group, as in 9. The ethyl analogue 9a turned
out to behave in the same way as its methyl counterpart (92%
ee, entry 32; compare with entry 3). Further extension of the
alkyl chain (with a terminal double bond, a possible site for
further elaboration), as in 9b-e, had only a marginal effect
on the enantioselectivity (84-95% ee, entries 33-36). Deri-
vatives 9f,g with an aromatic system at the terminus of the
alkyl chain exhibited one of the highest enantioselectivities
(95-96% ee, entries 37 and 38). Branching in the R2 group,
as in the isopropyl, cyclopropyl, and cyclobutyl derivatives
9h-j, had a positive effect (94-97% ee, entries 39-41). For
the bulkier cyclohexyl derivative 9k, where the two six-
membered rings differ in their electronic nature (Ph vs.
c-Hex), a minor decrease in enantioselectivity was observed
(76% ee, entry 42). On the other hand, reduction of the
highly congested tert-butyl derivative 9l proved very sluggish
and gave an almost racemic product (entry 43). Good
enantioselectivity was restored for the furyl isopropyl ana-
logue 9m (85% ee, entry 44). Finally, the case of the Ph/c-Hex
pattern (9k, entry 42) raised the question as to the role of the
aromatic ring, namely whether imines with two electroni-
cally different aromatic groups could also be reduced en-
antioselectively.22,23 To this end, we examined imine 9n, with
two essentially isosteric but electronically opposite aromatic
systems. The latter derivative turned out to be reduced
readily but the product was almost racemic (entry 45),
showing that the electronics in this imine reduction plays a
negligible role compared to steric effects.
An attempted switch from the CH2Cl group of the latter
set of imines to the analogous CH2CN derivatives created a
new situation owing to the π-system of the CtN: thus,
instead of the desired imines 10j,k, the corresponding con-
jugated enamines 31j,k were actually obtained (Scheme 2) on
reaction of the β-keto nitriles with anisidine. Nevertheless, in
the presence of a trace of a Brønsted acid, the enamines could
be equilibrated with the imine (although the enamine species
would still prevail in the mixture). After a series of experi-
ments, addition of acetic acid (1 equiv) was identified as a
healthy compromise between reactivity and selectivity: the
reduction, carried out in its presence, afforded the expected
β-amino nitriles 15j,k in high yields and enantioselectivities
(87% ee, entries 55 and 56). A single crystallization of 15j
from hexane furnished an enantiopure product (99.9% ee).
Reduction of the analogous enamines, generated from
β-keto esters (31l-n), afforded the expected β-amino esters
15l-n (again via the corresponding imines 10l-n, generated
in situ by the AcOH-catalyzed equilibration) also in high
yields and enantioselectivities (88-90% ee, entries 57-59).
A single crystallization of the β-amino ester 15l from hexane
increased its enantiopurity from 89% ee (entry 57) to 99.8%
ee. Finally, with the lower homologue 10o, lacking the CH2
group, the problem of enamine-imine dichotomy was elimi-
nated but the enantioselectivity proved to be rather low (59%
ee, entry 60). On the other hand, the higher homologue 10p
was reduced with high enantioselectivity again (88% ee,
entry 61).
A functional group in the side chain of the imine was
another feature to be explored as its presence would con-
siderably broaden the synthetic utility of this chemistry.
Therefore, the chloromethyl imines 10a-i were prepared
(vide supra) and investigated. Their reduction proceeded
very well and afforded the corresponding amines in >90%
ee (entries 46-54). The resulting amino chlorides 15a-i were
then cyclized to the corresponding aziridines on treatment
with t-BuOK.9d It is pertinent to note that the preparation
of the sensitive chloro imines 10a-i from the corresponding
R-chloro ketones was not entirely free of problems: while the
electron-neutral and electron-poor imines 10a-c were
synthesized and isolated as individual substances, their
electron-rich counterparts 10g,h could not be obtained as
pure compounds, since the reaction did not proceed to
completion. Therefore, in the latter instances, we chose to
Reduction of the cyclic imines 32a-d (Scheme 3), closely
related to the acyclic imines 6-10, was briefly investigated.
In spite of the familiar structural features, namely the N-aryl
substituent, the phenyl group in the position of the original
R1 substituent, and a functional group or an extended
carbon chain (R2), the reduction turned out to proceed with
much lower enantioselectivities than those observed for the
acyclic imines. Thus, 32a and 32b afforded the respective
amines 33a and 33b with 25% ee and 46% ee (entries 62 and
63), interestingly of opposite configuration to each other.24
The tetralone-derived imine 32c, which can be regarded as a
cyclic analogue of 9a, produced racemic amine 33c (entry 64)
and the indanone-derived imine 32d was reduced to amine
33d with 10% ee (entry 65). The latter results strongly suggest
(24) The absolute configuration of 33a,b was inferred by comparison of
the HPLC trace and optical rotation of our products with those obtained by
Rueping.8d The original assignment was made by Rueping by X-ray crystal-
lography for one of the members of the series and was then extrapolated to
the rest of the set.8d Hence, the configuration of 33a,b has not been rigorously
proven but is likely to be as shown in Scheme 3, i.e., (R)-(+)-33a and (S)-(-)-
33b.
(22) For an enantioselective reduction of isosteric, electronically biased
ketones, see: (a) Corey, E. J.; Helal, C. J. Tetrahedron Lett. 1995, 36, 9153. (b)
Corey, E. J.; Helal, C. J. Tetrahedron Lett. 1996, 37, 5675.
(23) For a review, see: Corey, E. J.; Helal, C. J. Angew. Chem., Int. Ed.
1998, 37, 1987.
5844 J. Org. Chem. Vol. 74, No. 16, 2009

Malkov et al.
JOCArticle
SCHEME 3. Asymmetric Reduction of Selected Cyclic Keti-
mines
(compare Table2, entry14with Table1, entry14andTable2,
entry 5). Finally, the sulfinamide 3013f has shown consis-
tently high enantioselectivities with imines derived from
acetophenone and its congeners (e96-97% ee; Table 2,
entry 15). However, this catalyst is noteworthy for its
efficiency with imines derived from nonaromatic amines,
such as PhCH2NH2; other N-alkyl imines exhibited rather
lower selectivities.13f Unfortunately, comparison with cata-
lyst 2513a is hampered by the fact that the authors did not
reveal the configuration of the resulting amine.
From the limited data reported by Matsumura,12 Sun,13
and Zhang,14 the pipecolinic acid-derived catalyst 25 and
the sulfinamide 30 appear to exhibit the highest enantios-
electivities, similar to those attained by us with Sigamide 23.
However, our substrate portfolio is now much broader
(Table 1) so that a more detailed comparison cannot be
made at present. Furthermore, Sigamide performed consis-
tently well at our standard 5 mol % loading and in selected
examples was shown to operate with the same efficiency even
when as little as 1 mol % had been used. Most of the catalysts
listed in Table 2 were used at 10-20 mol % loading, although
some cases with 5 mol % were also reported. Finally, our
reductions were carried out at room temperature, whereas
most experiments listed in Table 2 were run at 0 °C or below.
Mechanistic Considerations. Our experiments, in conjunc-
tion with those reported by Sun,13a,b,e,f indicate that the
Lewis base-catalyzed imine reduction with trichlorosilane
is not affected by isomeric nonhomogeneity of the starting
imines. Thus, for example, imines 9h-j, which exist as 5:2 to
5:3 (E/Z) mixtures, were reduced to the corresponding
amines with 94-97% ee (Table 1, entries 39-41). Appar-
ently, traces of HCl, naturally present in the moisture-
sensitive Cl3SiH, trigger an (E/Z) equilibration of imines
6-10, which is faster than the reduction (Scheme 4).
Brønsted acid-catalyzed isomerization of compounds con-
taining CdN bonds is a well-known process. The generally
accepted mechanism involves the initial protonation of the
imine nitrogen to generate trace amounts of the iminium ion
34, followed either by rotation about the C-N bond (35) or
via a nucleophilic attack by the acid counterion to produce
the corresponding tetrahedral intermediate, which under-
goes rotation/proton exchange on nitrogen and subsequent
elimination to give the other isomer.26 An analogous (Z/E)
isomerization mechanism, i.e., protonation at the R-position
to give the iminium ion 34, rotation around the C-C bond,
and proton elimination, can be envisaged in the case of
β-enamino nitriles and esters 31j-n. As discussed above,
the β-enamino esters 31l-n were obtained as the (Z)-isomers,
which are apparently more stable owing to an intramolecular
hydrogen bonding between the N-H and the ester carbonyl
group. In the case of 31l, no signs of isomerization in the
presence of acetic acid (2 equiv) could be detected by NMR
spectroscopy. By contrast, the β-enamino nitriles 31j,k read-
ily equilibrate to a mixture of (Z/E) isomers, e.g., 31j
isomerizes to a mixture of isomers (∼2.6:1) in CDCl3 solu-
tions overnight even without an external source of protons.9g
that highly enantioselective reduction occurs in a conforma-
tion that is attainable by the noncyclic imines 6-10 but
which is not available for the rigid cyclic structures, in
particular 32c,d.
Other Organocatalysts for the Reduction of Imines with
Trichlorosilane. To complete this study, we present a com-
parison of our findings with the highlights of the results
attained with catalysts 24-30 (Chart 2), as published by the
groups of Matsumura,12 Sun,13 and Zhang.14 These groups
have studied more limited sets of substrate imines and from
their work we have selected representative examples
(Table 2), which are further limited for the sake of space to
the imines derived from acetophenone (6a,b), its cyclohexyl
analogue (7c), the cinnamyl analogue (7a), and those with a
functional group (10l,o). Unfortunately, some of the authors
did not reveal the absolute configuration of their products,
which makes the comparison less straightforward.25
Clearly, the highest enantioselectivities were attained for
the acetophenone-derived imines 6a,b with the pipecolinic
acid-derived catalyst 25 (Table 2, entries 2 and 3), which
resonates with our results (Table 1, entries 3-5). The proline-
derived formamide 2412a exhibited rather lower enantios-
electivities (compare entry 1 in Table 2, with entry 1 in
Table 1). The pipecolinic acid-derived catalyst 2513a gave
slightly higher enantioselectivities with the cinnamyl and
cyclohexyl derivatives 7a and 7c0 (Table 2, entries 4 and 5)
than did our Sigamide (Table 1, entries 12, 13, and 14). On
the other hand, the level of asymmetric induction attained
with catalysts 26 and 2812b,13b,c (Table 2, entries 6-10) was
lower than that observed for Sigamide (Table 1, entries 3-5,
57, and 60). The R-picolinic amide 2914a approached
the selectivity attained with 23 and 25 in the case of the
acetophenone-derived imines 6a,b (Table 2, entries 11-13)
but was no match in the case of the cyclohexyl derivative 7c
(26) (a) Jennings, W. B.; Al-Showiman, S.; Tolley, M. S.; Boyd, D. R.
J. Chem. Soc., Perkin Trans. 2 1975, 1535. (b) Johnson, J. E.; Morales, N.
M.; Gorczyca, A. M.; Dolliver, D. D.; McAllister, M. A. J. Org. Chem. 2001, 66,
7979.During the preparation of this manuscript, a similar discussion of the role of
Brønsted acids on the reduction of imines with Hantzsch dihydropyridines
appeared: (c) Marcelli, T.; Hammar, P.; Himo, F. Chem.;Eur. J. 2008, 14, 8562.
(25) In some cases, where neither the absolute configuration nor the
optical rotations were stated in the paper, the configuration could be deduced
by comparison of the mobilities of our own samples on chiral HPLC columns
with those that could be extracted from the relevant Supporting Information.
J. Org. Chem. Vol. 74, No. 16, 2009 5845

Malkov et al.
JOCArticle
SCHEME 4. Imine (E/Z) and Imine/Enamine Equilibration
The key importance of H+ (at low concentrations) for the
reaction to occur suggests that protonated imines, i.e., the
corresponding iminium ions 34, might be the actual species
reduced by trichlorosilane. It can be speculated that the
protonated species 34 coordinates to the catalyst, e.g., via
hydrogen bonding to the anilide carbonyl group (36; Chart 3).
Trichlorosilane apparently is activated by coordination to
the formamide carbonyl group to form a pentacoordinated
silicon species. However, increasing the concentration of
Brønsted acid in the reaction mixture leads to an erosion
of enantioselectivity due to the competing nonselective
reduction (37). In the case of 7e and 9l, it is likely that steric
hindrance, exerted by the bulky t-Bu substituent (38), im-
pedes the protonation and/or coordination to the catalyst,
resulting in low reactivity and enantioselectivity. Compari-
son with the results obtained in the reduction of cyclic imines
32a-d (Table 1, entries 62-65) strongly suggests that highly
enantioselective reduction occurs in a conformation that is
attainable by the noncyclic imines 6-10 but which is not
available for the rigid cyclic structures (39). Molecular
modeling27 and the available crystallographic data for keti-
mines28 indicate that the N-aryl moiety adopts a nearly
perpendicular conformation with respect to the CdN bond,
whereas the aromatic R1 group is coplanar with it. It is likely
that in the transition state the iminium ion 34 adopts a
conformation where the R1 group is not coplanar with the
CdN bond to minimize the steric interaction with the
catalyst as well as the interaction of the ortho substituent
of the R1 group with the proton at the nitrogen. Clearly, such
a conformation cannot be attained by cyclic imines 32c,d,
where the aromatic R1 group is locked in the coplanar
arrangement 39. Furthermore, the reduced conformational
mobility of the N-aryl moiety, which is nearly coplanar with
the CdN bond in imines 32a,b, has a similar effect, i.e., the
reduction proceeds with much lower enantioselectivities
than those observed for the acyclic imines.
It is pertinent to note that the proline-derived catalyst 24,
which has the same absolute configuration as our Kenamide
(17) and Sigamide (23), induced the formation of (R)-11a,12a
i.e., of the opposite enantiomer to that produced by our
catalysts. The latter result suggests that these two types
of catalysts assume a different conformation on coordina-
tion of the silicon. This is partially supported by single crystal
X-ray analysis of Kenamide (17) and catalysts 40 and 41
(Chart 4).
The N-methyl moiety of catalyst 17 is crucial for the
enantioselectivity, presumably by controlling the spatial
orientation of the formamide group and relaying the stereo-
chemical information from the chiral center.9b The dihedral
angle CH3-N-CH-R, which reflects the magnitude of the
chiral relay, was found to be ca. 60° in Kenamide 17 in the
crystal (Table 4). For the tert-leucine-derived catalyst 40,
which showed a marginally reduced efficiency (e83% ee),9b
the corresponding dihedral angle CH3-N-CH-R was
∼88°. The crystals of the phenylalanine-derived catalyst 41,
exhibiting a further reduced level of asymmetric induction
(e49% ee),9b contain two crystallographically independent
Apparently, in the case of imines 6-8, and 9a-m, it is
the more stable (E) isomer that is reduced from the re-face
to give the (S) enantiomer. Accordingly, for chloro imines
10a-i the (Z) isomer is reduced from the enantiotopic si-face
to give the (R) enantiomer (note the change of Cahn-
Ingold-Prelog priorities here!). On the other hand, the
complete absence of steric preference for (E) or (Z) isomer
may account for the loss of enantioselectivity and formation
of a racemic product in the case of imine 9n.
We suspected that protonation would not only catalyze
the isomerization, but may also contribute to the nonselec-
tive background reaction by enhancing the electrophilicity of
the imines (via 34). Therefore the effect of a few acidic and
basic additives on the reduction of imines 6b and 10l,
catalyzed by Sigamide (23), was briefly investigated
(Table 3). Reduction of ketimine 6b in the presence of triflic
acid (0.1 equiv) led to a minor decrease of enantioselectivity
(91% ee, Table 3, entry 1; compare with 94% ee, Table 1,
entry 3); enantioselectivity was further reduced in the pre-
sence of a stoichiometric amount of methanol, which appar-
ently hydrolyzes Cl3SiH to produce HCl (Table 3, entry 2). In
the case of β-enamino nitriles and esters 31j-n (reacting as
their imine isomers 10j-n), addition of acetic acid (1 equiv)
was found to have a beneficial effect on reactivity at the slight
expense of enantioselectivity (vide supra).9g Increased
amounts of acetic acid in the reduction of 10l/31l also slightly
eroded the enantioselectivity (Table 3, entry 3; compare with
Table 1, entry 57), whereas the reduction, carried out in the
presence of the stronger trifluoroacetic acid, afforded a
completely racemic amine 15l (Table 3, entry 4).
By contrast, proton scavengers, such as i-Pr2EtN
€
(Hunig’s base) or 1,8-bis(dimethylamino)naphthalene
(proton sponge), were found to slow down the reaction
dramatically and to ruin the enantioselectivity (Table 3,
entries 5 and 6). An analogous effect of a stoichiometric
amount of 2,6-lutidine has been reported recently13g without
a detailed explanation. However, the beneficial effect of
substoichiometric amounts of the latter base on enantios-
electivity, reported by Sun,13g was not observed with our
model compounds (Table 3, entry 7; compare with Table 1,
entry 3).
The complete loss of reactivity in the presence of stoichio-
metric amounts of a base implies the involvement of a
Brønsted acid in the catalytic cycle. This hypothesis is sup-
ported by the observation that addition of acetic acid (1.5
equiv) to the reaction mixture, first “poisoned” with proton
sponge (1 equiv), completely restored the reactivity, providing
amine 11b in 92% yield and 87% ee (Table 3, entry 8).
(27) SPARTAN 004 for Windows; Wavefunction, Inc., 1840 Von Karman
Avenue, Suite 370, Irvine, CA 92612.
(28) For the X-ray structure of 32d, see ref 4c. For the X-ray structure of
3,5-dimethyl-N-[(E)-10-(40-nitrophenyl)ethylidene]aniline, see: Yang, X.-Y.;
Li, Y.-F.; Zheng, J.; Jian, F.-F. Acta Crystallogr. 2007, E63, No. o1611.
5846 J. Org. Chem. Vol. 74, No. 16, 2009

Malkov et al.
JOCArticle
TABLE 2. Examples of the Reduction of Selected Ketimines with Trichlorosilane, Catalyzed by Lewis Bases 24-30a
entry
catalyst (mol %)
imine
R1
R2
Ar
temp (°C)
solvent
yield (%)
imine (% ee)b
ref
1
2
3
4
5
6
7
8
24 (10)
25 (5)
6a
6a
6b
7a
7c0
6a
6a
6b
10l
10o
6a
6a
6b
7c
Ph
Ph
Ph
Me
Me
Me
Me
Me
Me
Me
Me
CH2CO2Et
CO2Me
Me
Me
Me
Ph
Ph
rt
0
0
0
0
-20
rt
rt
rt
rt
0
-10
-10
-10
0
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CHCl3
CHCl3
CHCl3
CHCl3
toluene
52
87
98
81
75
95
86
90
65
80
90
88
90
71
98
66 (R)
94
92
87
87
12a
13a
13a
13a
13a
13a
12b
12b
12b
12b
14a
14a
14a
14a
13f
25 (10)
25 (10)
25 (10)
26 (10)
28 (10)
28 (10)
28 (10)
28 (10)
29 (10)
29 (20)
29 (20)
29 (20)
30 (10)
4-MeO-C6H4
Ph
2-MeO-C6H4
Ph
Ph
4-MeO-C6H4
H
Ph
PhCHdCH
c-C6H11
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
89
73 (S)
75 (S)
41 (S)
45 (R)c
86 (R)
92 (R)
93 (R)d
61
9
10
11
12
13
14
15
Ph
Ph
4-MeO-C6H4
4-MeO-C6H4
CH2Ph
c-C6H11
Ph
Me
Me
6i
96 (R)
aThe reactions were carried out under similar condition as our own experiments (Table 1 and Experimental Section). bThe absolute configuration was
either stated by the authors or deduced by us (ref 25). cNote the change in the preference of the substituents in the Cahn-Ingold-Prelog system. dThe
configuration can be assumed to be (R) but direct evidence has not been provided.
TABLE 3. Reduction of Ketimines 6b and 10l with Trichlorosilane in the Presence of Additives, Catalyzed by Sigamide (S)-23 (5 mol %)a
entry imines 6-10 R1
R2
additive
amine 11b and 15l yield (%) 11b and 15l (% ee)
1
2
3
4
5
6
7
8
6b
6b
10l
10l
6b
6b
6b
6b
Ph Me
Ph Me
CF3SO3H (0.1 equiv)
MeOH (1 equiv)
11b
11b
15l
90
90
95
89
91
84
85
0
Ph CH2CO2Et CH3CO2H (1.5 equiv)
Ph CH2CO2Et CF3CO2H (1 equiv)
Ph Me
Ph Me
Ph Me
Ph Me
15l
i-Pr2EtN (1 equiv)
proton sponge (1 equiv)
2,6-lutidine (0.3 equiv)
11b
11b
11b
trace
0
73
0
-
92
87
proton sponge (1 equiv), then CH3CO2H (1.5 equiv) 11b
92b
aThe reaction was carried out at 0.2 mmol scale with 2.0 equiv of Cl3SiH at 18 °C for 16 h in toluene. bAcetic acid (1.5 equiv) was added after 10 min.
CHART 3
CHART 4
TABLE 4. Crystallographic Data for Amides 17 and 40-42
catalyst
dihedral angle CH3-N-C*H-R, deg
(S)-17a
(()-40
(S)-41b
(S)-42c
58.08, 62.75, 58.78
88.32
48.23, 67.33
7.66d
aThree crystallographically independent molecules in the unit cell.
c
bTwo crystallographically independent molecules in the unit cell. Re-
ference 29. dDihedral angle CH2-N-C*H-CH2.
species, which differ in the dihedral angle quite substantially
(∼48° and 67°). The X-ray structure of the proline-derived
catalyst 24 is not available for comparison; however, the
corresponding dihedral angle CH2-N-CH-CH2 in the
structurally related acetamide 42 was reported to be much
smaller (7.66°).29 Although the solid-state data may not be
directly used for the interpretation of the behavior of mole-
cules in a solution, the dramatically different conformation
of the flexible catalysts 17, 40, and 41 vs. that of the rigid 42
can be assumed to be reflected, to some extent, in the
solution. Hence, the reversed sense of asymmetric induction
for those two types of catalysts can be attributed to the
constrained character of 24 as opposed to the flexibility of
catalysts 17 and 40, which apparently prefer a different
conformation. The flexibility of the valine-derived catalysts,
combined with the chiral relay effect of the N-Me group,
presumably allows an optimal conformation in the transition
state. The valine-derived catalysts (S)-17 and (S)-23 and the
proline-derived analogue (S)-2412a thus represent the two
(29) Matsuzaki, T.; Iitaka, Y. Acta Crystallogr. 1971, B27, 507.
J. Org. Chem. Vol. 74, No. 16, 2009 5847

Malkov et al.
JOCArticle
extremes, favoring the formation of (S)-11a (e92% ee) and
(R)-11a (e66% ee), respectively.
9m,36,37 9b-e and 9h,37 and 9n38 were synthesized according to
the published procedures. In this section, only the general
procedures are shown. The analytical data for the imines
and amines and copies of their NMR spectra and HPLC traces
are given in the Supporting Information; some of these are
known3g,p,8d,9g,39-47 but were often prepared by different
procedures.
General Procedure for the Preparation of Imines: Method A.
Molecular sieves (5 A; 6.25 g) were added to a solution of the
ketone (5.00 mmol; 1 equiv) and p-anisidine (770 mg, 6.25 mmol;
1.25 equiv) in anhydrous toluene (25 mL) and the reaction
mixture was heated under reflux for 5 h. The mixture was then
cooled, the sieves were filtered off, the filtrate was evaporated,
and the residue was purified by flash chromatography on a silica
gel column (pretreated overnight with 10% triethylamine in
petroleum ether) with a petroleum ether-ethyl acetate mixture
(100:0 to 60:40).
Conclusions
We have demonstrated the broad scope of the reduction of
ketimines with trichlorosilane catalyzed by the Lewis-basic
Sigamide 23 (Scheme 1 and Table 1). High enantioselectivity
(typically g90% ee) was observed across the spectrum of
aromatic, heteroaromatic, and aliphatic substrates, which
may contain additional functional groups. The reaction
proceeds in toluene at room temperature overnight with
1-5 mol % of the catalyst. Hence, our protocol compares
favorably with its alternatives, such as catalytic hydrogena-
tion (which requires high-pressure equipment in the case
of imines),3,6 or reduction with Hantzsch dihydropyridine
catalyzed by chiral acids.8 Furthermore, our catalytic pro-
cess is complementary to the Cu-catalyzed hydrosilylation
developed by Lipshutz,4f-h which favors a 1,4-addition to
R,β-unsaturated imines, whereas our system is 1,2-selective
(Table 1, entries 12 and 13). Current limitations are relatively
few: (1) the reaction exhibits very low enantioselectivity with
imines derived from pyridine (Table 1, entries 17 and 18) but
a remedy to this flaw was found (Table 1, entry 19); (2)
reduction of ketimines derived from diaryl ketones gives
practically racemic products even if the two aryl groups
differ in their electronics (Table 1, entry 45); (3) the current
system only works efficiently with imines derived from
aromatic amines (e.g., aniline and anisidine),30 nevertheless,
the anisidine-derived amines can be oxidatively deprotected
to produce primary amines;9g and (4) imines derived from
cyclic ketones exhibit low enantioselectivity. Some of these
limitation are now being addressed in this Laboratory and
the results will be communicated in due course.
Method B. A solution of the ketone (5.0 mmol; 1 equiv),
p-anisidine (647 mg, 5.25 mmol, 1.05 equiv), and p-toluenesul-
fonic acid (47 mg, 0.25 mmol, 5 mol %) in anhydrous benzene
(25 mL) was heated under reflux with a Dean-Stark trap for
10 h, then cooled and evaporated to dryness. The residue was
purified by flash chromatography on a silica gel column (treated
overnight with 10% triethylamine in petroleum ether) with a
petroleum ether-ethyl acetate mixture (100:0 to 90:10).
Method C18. A 1.0 M solution of titanium(IV) chloride in
CH2Cl2 (3.0 mL, 3.0 mmol, 1 equiv) was added dropwise to a
precooled (0 °C) solution of the ketone (3.0 mmol, 1 equiv)
and p-anisidine (1.11 g, 9.0 mmol, 3 equiv) in anhydrous ether
and the reaction mixture was heated to reflux overnight. The
mixture was then cooled to room temperature and filtered, then
the solid material was washed with anhydrous ether. The filtrate
was then washed with a KOH solution (2 M, 15 mL) and the
aqueous layer was extracted with ether (3 ꢀ 10 mL). The
combined organic layers were dried over K2CO3 and evapo-
rated. The crude mixture was purified on a silica gel column
(30 mL) (treated overnight with 10% triethylamine in petroleum
ether) in a petroleum ether-Et3N mixture (99:1).
Trichlorosilane, a nonexpensive stoichiometric reducing
agent, is relatively easy to handle under anhydrous (but not
necessarily anaerobic) conditions. The aqueous workup
produces NaCl and SiO2, two benign inorganics, which in
conjunction with the use of toluene as the optimal solvent
renders this method environmentally acceptable.
Method D9g,20. A solution of the corresponding ketonitrile
(5.0 mmol, 1 equiv) and p-anisidine (800 mg, 6.5 mmol, 1.3
equiv) or aniline (605 mg, 6.5 mmol, 1.3 equiv) in glacial acetic
acid (2.9 mL) was stirred at 80 °C for 6 h under an argon
atmosphere. The reaction mixture was then cooled to room
temperature and the precipitated solid was filtered off, washed
with glacial acetic acid, vacuum dried, and used without further
purification (31j) or after an additional recrystallization from
aqueous methanol (31k).
Experimental Section
The commercially unavailable ketones required for the pre-
paration of imines 7b,31 8c,32 8e,33 8j and 8k,34 8n,35,36 8m and
Method E21. A solution of the corresponding ketoester (5.0
mmol, 1 equiv), p-anisidine (677 mg, 5.5 mmol, 1.1 equiv), and p-
toluenesulfonic acid (95 mg, 0.5 mmol, 10 mol %) in dry ethanol
(5 mL) was refluxed for 24-48 h under an argon atmosphere,
then cooled and evaporated to dryness. Solid residues were
directly purified by crystallization. Oily residues were dissolved
in dichloromethane (20 mL) and washed with water (10 mL).
The organic phase was dried (MgSO4), filtered, and evaporated,
(30) This limitation seems to have been eliminated by the sulfinamide
catalyst 30 and partly by the R-picolinamide 29.
(31) Kawai, Y.; Saitou, K.; Hida, K.; Dao, D. H.; Ohno, A. Bull. Chem.
Soc. Jpn. 1996, 69, 2633.
(32) Fontana, F.; Minisci, F.; Nogueira Barbosa, M. C.; Vismara, E. Acta
Chem. Scand. 1989, 43, 995.
(33) (a) McMorris, T. C.; Chavez, R. G.; Patil, P. A. J. Chem. Soc., Perkin
Trans. 1 1996, 295. (b) Birkinshaw, T. N.; Harkin, S. A.; Kaye, P. T.; Meakins,
G. D.; Smith, A. K. J. Chem. Soc., Perkin Trans.1 1982, 939. (c) Ferreira, L. M.;
Lobo, A. N.; Prabhakar, S. Tetrahedron 1999, 55, 3541.
(39) Ahlbrecht, H.; Fischer, S. Tetrahedron 1970, 26, 2837.
(40) Liu, S.; Yu, Y.; Liebeskind, L. S. Org. Lett. 2007, 9, 1947.
(41) Kato, H.; Wakao, K.; Yamada, A.; Mutoh, Y. J. Chem. Soc., Perkin
Trans. 1 1988, 189.
(42) Roelants, F.; Bruylants, A. Tetrahedron 1978, 34, 2229.
(43) Staskun, B.; Israelstam, S. S. J. Org. Chem. 1961, 26, 3191.
(44) Itoh, T.; Nagata, K.; Miyazaki, M.; Ishikawa, H.; Kurihara, A.;
Ohsawa, A. Tetrahedron 2004, 60, 6649.
ꢀ
(34) (a) Lukacs, G.; Porcs-Makkay, M.; Simig, G. Eur. J. Org. Chem. 2004,
4130. (b) Domagala, J. M. Tetrahedron Lett. 1980, 21, 4997. (c) Kim, S.; Lee,
J. I.; Kim, Y. C. J. Org. Chem. 1985, 50, 560. (d) Sterzycki, R. Synthesis 1979,
724.
(35) Kaufman, K. D.; Erb, D. J.; Blok, T. M.; Carlson, R. W. J. Heterocycl.
Chem. 1982, 19, 1051.
(36) Corey, E. J.; Schmidt, G. Tetrahedron Lett. 1979, 399.
ꢀ
(37) Hideg, K.; Hankovszky, H. O.; Halasz, H. A. J. Chem. Soc., Perkin
(45) Menche, D.; Hassfeld, J.; Li, J.; Menche, G.; Ritter, A.; Rudolph, S.
Org. Lett. 2006, 8, 741.
(46) Canella, R.; Clerici, A.; Pastori, N.; Regolini, E.; Porta, O. Org. Lett.
2005, 7, 645.
(47) Hasegawa, M.; Taniyama, D.; Tomioka, K. Tetrahedron 2000, 56,
10153.
Trans. I 1988, 2905.(b) Frenette, R.; Monette, M.; Bernstein, M. A.; Young,
R. N.; Verhoeven, T. R. J. Org. Chem. 1991, 56, 3083. (c) Matsuya, Y.;
Sasaki, K.; Nagaoka, M.; Kakuda, H.; Toyooka, N.; Imanishi, N.; Ochiai,
H.; Nemoto, H. J. Org. Chem. 2004, 69, 7989.
(38) Wilkinson, M. C.; Saez, F.; Hon, W. L. E. Synlett 2006, 1063.
5848 J. Org. Chem. Vol. 74, No. 16, 2009

Malkov et al.
JOCArticle
then the residue was purified by column chromatography on
silica gel.
114.77(2 ꢀ CH), 126.46 (2 ꢀ CH), 127.21 (CH), 128.71(2 ꢀ CH),
141.39 (C), 143.33 (C), 151.97 (C), 174.07 (C); IR ν 3394, 3025,
3005, 2951, 2833, 1733, 1513, 1452, 1237 cm-1; MS (EI)
m/z (%) 299 (M+•, 30), 212 (100), 117 (23), 85 (52), 83 (80);
HRMS (EI) 299.1518 (C18H21NO3 requires 299.1521); HPLC
analysis [Chiralpak IB, hexane-propan-2-ol (90:10), 0.75 mL/
min, tminor=16.52 min, tmajor=28.60 min] showed 88% ee.
2-Methoxy-1-(40-methoxyphenyl)-5-phenyl-1H-pyrrole. The
title compound was obtained as white crystals, isolated in
27% yield as a less polar component along with 15p: mp 101-
Methyl (E)-4-(N-40-Methoxyphenylimino)-4-phenylbutanoate
(10p). Method B (60%; chromatographic purification with a
97:3 mixture of petroleum ether and AcOEt): yellow crystals; mp
62-63 °C (hexane); 1H NMR (400 MHz, CDCl3, a mixture of
(E/Z) isomers in ∼7:3 ratio; the minor one is marked with an *)
δ 2.42-2.47 (m, 2H), 2.81* (dd, J=7.0, 6.9 Hz, 0.86H), 3.00-
3.07 (m, 2.86H), 3.60 (s, 3H), 3.698* (s, 1.29H), 3.701* (s,
1.29H), 3.82 (s, 3H), 6.51-6.55* (m, 0.86H), 6.63-6.67* (m,
0.86H), 6.71-6.75 (m, 2H), 6.89-6.92 (m, 2H), 7.98-7.12* (m,
0.86H), 7.20-7.24* (m, 1.29H), 7.42-7.47 (m, 3H), 7.85-7.90
(m, 2H); 13C NMR δ 25.18 (CH2), 30.30* (CH2), 31.99 (CH2),
35.50* (CH2), 51.67* (CH3), 51.86 (CH3), 55.28* (CH3), 55.47
(CH3), 113.77* (2 ꢀ CH), 114.45 (2 ꢀ CH), 120.08 (2 ꢀ CH),
122.12* (2 ꢀ CH), 127.49 (2 ꢀ CH), 127.78* (2 ꢀ CH), 128.20*
(2 ꢀ CH), 128.47* (CH), 128.66 (2 ꢀ CH), 137.91 (C), 137.99*
(C), 143.74* (C), 144.31 (C), 155.69* (C), 156.00 (C), 168.29 (C),
169.20* (C), 172.40 (C), 173.80* (C); IR ν 3020, 2951, 2834,
11735, 1630, 1503, 1439, 1287, 1241 cm-1; MS (CI/isobutane)
m/z (%) 298 [(M + H)+, 100]; HRMS (CI/isobutane) 298.1441
(C18H20NO3 requires 298.1443).
1
102 °C (hexane); H NMR (400 MHz, CDCl3) δ 3.80 (s, 3H),
3.81 (s, 3H), 5.45 (d, J=3.7 Hz, 1H), 6.26 (d, J=3.7 Hz, 1H),
6.84-6.88 (m, 2H), 7.03-7.16 (m, 7H); 13C NMR δ 54.30
(CH3), 56.71 (CH3), 82.86 (CH), 105.73 (CH), 112.92 (2 ꢀ
CH), 124.24 (CH), 125.80 (C), 126.13 (2 ꢀ CH), 126.95 (2 ꢀ
CH), 128.10 (2 ꢀ CH), 129.00 (C), 132.24 (C), 149.12 (C), 157.39
(C); IR ν 3019, 2956, 2931, 2832, 1599, 1564, 1515, 1454, 1420,
1294, 1250, 1216 cm-1; MS (EI) m/z (%) 279 (M+•, 55), 264
(100), 193 (10), 103 (10), 77 (16); HRMS (EI) 279.1258
(C18H17NO2 requires 279.1259).
Acknowledgment. We thank the EPSRC for grant No.
GR/S87294/01, WestChem for a graduate fellowship to K.
V., Drs. Andy Parkin and Louis Farrugia for the X-ray
analysis, and Prof. Magnus Rueping of the Johann Wolf-
€
gang Goethe-Universitat Frankfurt for a generous donation
of samples of 32a,b.
General Procedure for Enantioselective Reduction of Imines.
Trichlorosilane (40 μL, 0.4 mmol, 2 equiv) was added drop-
wise to a cooled solution (0 °C) of the imine (0.2 mmol, 1 equiv)
and catalyst (0.01 mmol, 0.05 equiv) in anhydrous toluene
(2 mL) under an argon atmosphere and the reaction mixture
was allowed to stir at room temperature (15-20 °C) for 24 h
(unless otherwise stated). Then the reaction mixture was diluted
with ethyl acetate (5 mL) and quenched with a saturated
NaHCO3 solution (20 mL) and the layers were separated.
The aqueous layer was extracted with AcOEt (2 ꢀ 5 mL), the
combined organic phase was washed with water (2 ꢀ 15 mL)
and brine (5 mL), dried over anhydrous MgSO4, and evapo-
rated. The residue was purified by flash chromatography
on a silica gel column (25 mL) with a petroleum ether-ethyl
acetate mixture (98:2 to 85:15). The yields and enantioselec-
tivities are given in Table 1. The absolute configuration of the
amines was established by comparison of their optical rotation
(measured in CHCl3) with the literature data and/or by compari-
son of their HPLC behavior with that of the authentic samples.
(-)-Methyl 4-[N-(40-Methoxyphenyl)amino]-4-phenylbutano-
ate (15p). 15p was obtained as a yellow oil, isolated as a more
polar component from the reduction of 10p: mp 59-60 °C
(hexane/CH2Cl2); [R]D -13.5 (c 1.0, CHCl3); 1H NMR (400
MHz, CDCl3) δ 2.03-2.19 (m, 2H), 2.41 (dd, J=7.3, 7.2 Hz,
3H), 3.66 (s, 3H), 3.69 (s, 3H), 3.98 (br s, 1H), 4.30 (dd, J=7.2,
6.5 Hz, 1H), 6.45-6.49 (m, 2H), 6.66-6.70 (m, 2H), 7.21-7.25
(m, 1H), 7.29-7.34 (m, 4H); 13C NMR δ 31.06 (CH2), 33.32
(CH2), 51.77 (CH3), 55.75 (CH3), 58.52 (CH), 114.60 (2 ꢀ CH),
Note Added in Proof. (1) Coordination of pyridine to the
silicon atom of Cl2SiHR, relevant to the behavior of the pyridine
derivatives 8a-c (Table 1, entries 17-19), has been demon-
strated by X-ray crystallography and solution studies: Fester, G.
€
W.; Wagler, J.; Brendler, E.; Bohme, U.; Gerlach, D.; Kroke, E.
J. Am. Chem. Soc. 2009, 131, 6855. (2) A recent mechanistic and
computational study on the reduction of imines with Cl3SiH,
catalyzed by a series of amides, including some of those dis-
cussed in this paper, suggests that the catalyst not only coordi-
nates to the reagent but also acts as a proton donor to the
imine in the transition structure: Zhang, Z.; Rooshenas, R.;
Hausmann, H.; Schreiner, P. R. Synthesis 2009, 1531. However,
this mechanistic picture does not take into account the role of
the added Brønsted acid observed in our experiments.
Supporting Information Available: Experimental methods,
1H and 13C NMR spectra for new compounds, crystallographic
data for 17 and 40-41 in CIF format, and HPLC traces for
chiral amines. This material is available free of charge via the
Internet at http://pubs.acs.org.
J. Org. Chem. Vol. 74, No. 16, 2009 5849