Structural effects on the catalytic, emulsifying, and recycling properties of chiral amphiphilic dendritic organocatalysts

Article abstract of DOI:10.1021/jo9006128

(Chemical Equation Presented) Three series of chiral amphiphilic G1-G3 dendritic organocatalysts containing an optically active polar proline-derived core and one or two nonpolar hydrocarbon dendrons were prepared. These dendritic organocatalysts were emp

Full text of DOI:10.1021/jo9006128

pubs.acs.org/joc
Structural Effects on the Catalytic, Emulsifying, and Recycling Properties
of Chiral Amphiphilic Dendritic Organocatalysts
Chui-Man Lo and Hak-Fun Chow*
Department of Chemistry and The Center of Novel Functional Molecules, The Chinese University of
Hong Kong, Shatin, NT, Hong Kong SAR
hfchow@cuhk.edu.hk
Received March 23, 2009
Three series of chiral amphiphilic G1-G3 dendritic organocatalysts containing an optically active polar
proline-derived core and one or two nonpolar hydrocarbon dendrons were prepared. These dendritic
organocatalysts were employed in the asymmetric aldol and nitro-Michael additions in oil-in-water
emulsions to reveal the effects of dendron size and branching on the catalytic properties. The incorporation
of larger hydrophobic dendrons has the advantages of promoting emulsion formation in water, improving
the reaction enantioselectivity, decreasing catalyst loading (to 1 mol %), and facilitating catalyst recovery
after the reactions. In general, the larger dendrons tended to lower catalyst reactivity due to their increasing
steric blocking effect. However, some astonishing observations were found in some of the G1 and G2
dendritic organocatalysts, wherein an increase in the steric bulkiness and branching of the dendron resulted
in better catalyst reactivity. It was also found that higher product yields and enantioselectivities were
obtained in the aldol reactions when the aromatic aldehyde contains an electron-withdrawing substituent.
The catalysts could be recycled and reused five times without significant drop in product yields and
enantioselectivities. In addition, cross product contamination was not found when the recovered G3
catalyst was subsequently used in another reaction involving different substrates.
Introduction
Proline and its derivatives are most commonly employed
These reactions involve the formation of an enamine be-
tween the secondary amino group in proline with the ketone
substrate and protonation of the aldehyde/imine reaction
partner via the carboxylic acid group of the proline moiety.
The stereochemical outcome of the reaction can be rationa-
1
as catalysts in the asymmetric aldol and Michael reactions.
2
*To whom correspondence should be addressed.
1) (a) Braun, M.; Ghosh, A. K.; Shevlin, M. Mukaiyama, T.; Matsuo,
3
lized by the Zimmerman-Traxler transition state model.
(
J.-i.; List, B.; Fessner, W.-D.; Tanaka, F.; Barbas, C. F., III; Schinzer, D.
Modern Aldol Reactions. In Enolates, Organocatalysis, Biocatalysis and
Natural Product Synthesis; Mahrwald, R., Ed.; Wiley-VCH: Weinheim, Ger-
many, 2004; Vol. 1. (b) Machajewski, T. D.; Wong, C. H. Angew. Chem., Int. Ed.
(2) (a) List, B.; Pojarliev, P.; Martin, H. J. Org. Lett. 2001, 3, 2423–2425.
(b) Enders, D.; Seki, A. Synlett 2002, 26–28. (c) Pansare, S. V.; Pandya, K. J.
Am. Chem. Soc. 2006, 128, 9624–9625. (d) Vicario, J. L.; Badıa, D.; Carrillo,
´
L. Synthesis 2007, 2065–2092. (e) Alma s- i, D.; Alonso, D. A.; N ꢀa jera, C.
Tetrahedron: Asymmetry 2007, 18, 299–365. (f) Chen, H.; Wang, Y.; Wei, S.;
Sun, J. Tetrahedron: Asymmetry 2007, 18, 1308–1312. (g) Yan, Z.-Y.; Niu,
Y.-N.; Wei, H.-L.; Wu, L.-Y.; Zhao, Y.-B.; Liang, Y.-M. Tetrahedron:
Asymmetry 2007, 18, 3288–3293. (h) Zhu, S.; Yu, S.; Ma, D. Angew. Chem.,
Int. Ed. 2008, 47, 545–548.
2
000, 39, 1352–1374. (c) List, B.; Lerner, R. A.; Barbas, C. F. III J. Am. Chem.
Soc. 2000, 122, 2395–2396. (d) Sakthivel, K.; Notz, W.; Bui, T.; Barbas, C. F. III
J. Am. Chem. Soc. 2001, 123, 5260–5267. (e) List, B.; Pojarliev, P.; Castello, C.
Org. Lett. 2001, 3, 573–575. (f) Northrup, A. B.; MacMillan, D. W. C. J. Am.
Chem. Soc. 2002, 124, 6798–6799. (g) List, B. Tetrahedron 2002, 58, 5573–
5
590. (h) Tang, Z.; Yang, Z.-H.; Chen, X.-H.; Cun, L.-F.; Mi, A.-Q.; Jiang, Y.-Z.;
Gong, L.-Z. J. Am. Chem. Soc. 2005, 127, 9285–9289. (i) Guillena, G.; N ꢀa jera,
C.; Ram ꢀo n, D. J. Tetrahedron: Asymmetry 2007, 18, 2249–2293.
(3) Zimmerman, H. E.; Traxler, M. D. J. Am. Chem. Soc. 1957, 79, 1920–
1924.
DOI: 10.1021/jo9006128
r 2009 American Chemical Society
Published on Web 05/28/2009
J. Org. Chem. 2009, 74, 5181–5191 5181

Lo and Chow
JOCArticle
SCHEME 1. Aldol Reaction Promoted by Amphiphilic Organo-
catalysts 1
8
water due to major contributions by Hayashi, Barbas, and
9
1
0
C oꢀ rdova. One of the promising results reported by Barbas
involved the use of surfactant-type amphiphilic organocata-
lysts (e.g., 1) to give high product yields and good ee values
FIGURE 1. Organic transformation inside an emulsion system.
(
87-99%) in the asymmetric aldol reactions between ke-
11
tones and aldehydes in water (Scheme 1). As a result of
incorporating the surfactant appendage into the catalyst,
such amphiphilic diamino organocatalysts obviated the
necessity of adding an external surfactant into the reaction
medium. Because the organocatalysts do not contain acidic
functionalities, trifluoroacetic acid (TFA, 1 equiv) was
added as the proton source to enhance the electrophilicity
of the aldehyde substrate. The two long alkyl chains in
the catalyst then acted as the surfactant parts to promote
the formation of emulsion in water and to enhance
the reactivity and selectivity of the asymmetric aldol reac-
tions. In addition, the emulsion system could also over-
come mixing problems due to incompatible solubility
properties between the reactants. Hence, formation of a
steady emulsion is of ultimate importance to ensure high
product yields and good product enantioselectivities.
Despite these promising findings, there are little studies
on the effect of the size and shape of the surfactant appen-
dages on the micelle formation efficiency and the catalyst
properties.
In this paper, we report (a) the synthesis of several series of
G1-G3 chiral amphiphilic dendritic organocatalysts Di-L-
Gn (doubly dendronized, longer spacer), Mono-L-Gn (singly
dendronized, longer spacer), and Mono-S-Gn (singly den-
dronized, shorter spacer) that contain a proline catalytic core
with one or two nonpolar hydrocarbon (HC) dendron(s) on
the dendrimer surface, (b) their applications in asymmetric
aldol reactions and nitro-Michael additions, (c) the effect of
dendron size (i.e., G1 vs G2 vs G3), degree of branching (i.e.,
Di vs Mono), and spacer chain length (i.e., L vs S) between
the proline catalytic core and the dendron on the reaction
reactivity and enantioselectivity, and (d) their catalyst re-
covery property and reusability in different aldol reactions
without product cross contaminations. Our studies reveal a
positive branching effect of the nonpolar HC dendrons on
emulsion formation and catalyst reactivity, and provide the
first assessment of the size and geometry of the dendritic
appendages on the efficiency of emulsion formation. By
incorporating the catalytic center inside a dendritic skeleton,
recovery and recycling of the dendritic organocatalysts
become feasible. Furthermore, the branching nature of the
dendrons can promote better stereodifferentiation to give
higher product enantioselectivity. Although applications of
dendritic catalysts are well documented in the literature,
In general, good to excellent product enantioselectivities
(ee = 60-96%) could be achieved with electronically sui-
tably substituted substrates. However, two major drawbacks
of the reactions are the requirements of high catalyst loading
(>20 mol %) and excess amount of ketone donor to compete
against the undesirable decarboxylative side reaction of
4
proline. In view of these shortcomings, proline derivatives
with an alternative type of proton donor group, such as the
5
conjugated acid of a tertiary amine, were designed. Such
modified proline catalysts generally possess better solubility
in organic solvents and, hence, a lower catalyst loading can
be employed.
To further improve product yields and enantioselectivities
of the reactions, several research groups began to utilize
the hydrophobic effect of the catalyst system in water or in
1
2
6
an emulsion system. In the presence of water, the amphi-
philic organocatalyst will form a micelle structure, exposing
the hydrophilic sites to water, while the hydrophobic parts
will pack together to form a nonpolar pocket (Figure 1).
Once the organic reactants are added, they tend to penetrate
into the hydrophobic pocket. As a result, the reaction
proceeds in a concentrated organic phase generated by the
organic substrates and the organocatalyst inside the emul-
sion droplets. Barbas reported the first use of (S)-proline
in the presence of sodium dodecyl sulfate as a surfactant
to catalyze the aldol reaction in aqueous phosphate
buffer (pH = 7.4) to give aldol products in moderate to high
7
reaction yields (50-99%). However, the ee values were
poor, as only racemates were obtained. Developments
in this field progressed rapidly afterward and there were
significant breakthroughs in asymmetric organocatalysis in
(
4) Orsini, F.; Pelizzoni, F.; Forte, M.; Destro, R.; Gariboldi, P. Tetra-
hedron 1988, 44, 519–541.
5) (a) Mase, N.; Tanaka, F.; Barbas, C. F. III Angew. Chem., Int. Ed.
004, 43, 2420–2423. (b) Hayashi, Y.; Tamura, T.; Shoji, M. Adv. Synth.
(
2
Catal. 2004, 346, 1106–1110. (c) Andrey, O.; Alexakis, A.; Tomassini, A.;
Bernardinelli, G. Adv. Synth. Catal. 2004, 346, 1147–1168.
(
6) (a) Lindstr o€ m, U. M. Chem. Rev. 2002, 102, 2751–2772. (b) Manabe,
K.; Kobayashi, S. Chem.;Eur. J. 2002, 8, 4094–4101. (c) Chanda, A.; Fokin,
V. V. Chem. Rev. 2009, 109, 725–748.
(
7) C oꢀ rdova, A.; Notz, W.; Barbas, C. F. III Chem. Commun. 2002, 3024–
025.
8) (a) Hayashi, Y.; Sumiya, T.; Takahashi, J.; Gotoh, H.; Urushima, T.;
3
(
Shoji, M. Angew. Chem., Int. Ed. 2006, 45, 958–961. (b) Hayashi, Y.;
Aratake, S.; Okano, T.; Takahashi, J.; Sumiya, T.; Shoji, M. Angew. Chem.,
Int. Ed. 2006, 45, 5527–5529. (c) Hayashi, Y.; Urushima, T.; Aratake, S.;
Okano, T.; Obi, K. Org. Lett. 2008, 10, 21–24.
(
9) (a) C oꢀ rdova, A.; Barbas, C. F. III Tetrahedron Lett. 2003, 44, 1923–
926. (b) Mase, N.; Watanabe, K.; Yoda, H.; Takabe, K.; Tanaka, F.;
Barbas, C. F. III J. Am. Chem. Soc. 2006, 128, 4966–4967.
10) (a) Ibrahem, I.; C oꢀ rdova, A. Tetrahedron Lett. 2005, 46, 3363–3367.
b) C oꢀ rdova, A.; Zou, W.; Ibrahem, I.; Reyes, E.; Engqvist, M.; Liao, W.-W.
Chem. Commun. 2005, 3586–3588.
1
(11) Mase, N.; Nakai, Y.; Ohara, N.; Yoda, H.; Takabe, K.; Tanaka, F.;
Barbas, C. F. III J. Am. Chem. Soc. 2006, 128, 734–735.
(12) (a) H €a ger, M.; Currie, F.; Holmberg, K. Top. Curr. Chem. 2003, 227,
53–74. (b) Holmberg, K. Eur. J. Org. Chem. 2007, 731–742.
(
(
5
182 J. Org. Chem. Vol. 74, No. 15, 2009

Lo and Chow
JOCArticle
however, amphiphilic dendritic organocatalysts worked on
an emulsion system have not been disclosed before.
SCHEME 2. Retrosynthetic Analysis of Amphiphilic Dendritic
Organocatalysts
HC sector via the same three-carbon spacer unit (L) to the
proline core. Finally, the Mono-S-Gn series denote com-
pounds that employ a relatively short one carbon spacer (S)
linking one single dendronized HC sector and the proline
core. The differences in the number of dendron(s) and the
spatial distance between dendron(s) and catalytic core
should allow one to compare the effect of steric shielding
on the reactivity, enantioselectivity, and other properties of
the catalyzed reactions.
Synthesis. The preparation of doubly dendronized
Di-L-Gn catalysts is best achieved via a di-N-alkylation of
1
8
a proline-derived amine 2 to a HC dendron 3 bearing a
good leaving group (Scheme 2). The singly dendronized
catalysts, Mono-L-Gn and Mono-S-Gn, could be prepared
by a reductive amination starting from the same amine 2 to a
HC dendron 4 bearing an aldehyde focal point functionality.
The elongated G1-G3 HC dendritic bromides 5-7 were
prepared from the known R,β-unsaturated esters 8-10
1
7
(
Scheme 3). Hydrogenation of compounds 8-10 gave the
Results and Discussion
corresponding saturated esters 11-13 in 94-99% yield.
Subsequent reduction with LiAlH in THF at 0 °C gave
4
Catalyst Design. Due to the amphiphilic nature of the
catalyst, the dendrons used in the design must be highly
hydrophobic in nature. Previous reports showed that even
the corresponding alcohols 14-16 in 90-99% yield. The
target bromides 5-7 could be obtained from bromination
1
3
of 14-16, respectively, with NBS and PPh
2-97% yield (details of the synthesis of all intermediates
can be found in Supporting Information).
in CH Cl in
3 2 2
oligoether dendrons were highly polar, hence, only HC
dendrons were suitable candidates. Due to the high crys-
9
1
tallinity of oligo(phenylene), oligo(phenylenevinylene),
4
15
1
6
and oligo(phenyleneethynylene) dendrons, they are less
For the preparation of Di-L-G1 and Di-L-G2 catalysts,
di-N-alkylation of the proline-derived compound 2 with
the elongated G1 5 and G2 dendritic bromide 6 in the
presence of K CO , KI, 18-crown-6, and i-Pr NEt in DMF
at 60 °C afforded the Boc-protected dialkylation product
Boc-Di-L-G1 and Boc-Di-L-G2 in 74 and 77% yields, re-
spectively. After removal of Boc group by TFA, the target
dendritic catalysts, Di-L-G1 and Di-L-G2, were obtained
in 99% in both cases. However, for the synthesis of Di-L-G3
dendritic catalyst, the di-N-alkylation reaction was very
sluggish, possibly due to the highly sterically hindered en-
vironment caused by the larger dendrons. The resulting
di-N-alkylation product Boc-Di-L-G3 was obtained only in
1
7
desirable than aliphatic HC dendrons to be used in this
study. To explore the structural effects on the reactivity,
enantioselectivity, emulsion formation and recovery proper-
ties of the catalysts, three classes of amphiphilic dendritic HC
organocatalysts were designed. The Di-L-Gn series are com-
pounds bearing two (Di) dendronized HC sectors with a
relatively longer three-carbon spacer (L) connecting the
proline and two HC dendrons. The Mono-L-Gn series
feature the attachment of one single (Mono) dendronized
2
3
2
(
13) Hawker, C. J.; Wooley, K. L.; Fr ꢀe chet, J. M. J. J. Am. Chem. Soc.
993, 115, 4375–4376.
14) (a) Miller, T. M.; Neenan, T. X.; Zayas, R.; Bair, H. E. J. Am. Chem.
1
(
3
7% yield. At a higher reaction temperature (100 °C),
Soc. 1992, 114, 1018–1025. (b) Morgenroth, F.; Reuther, E.; M u€ llen, K.
Angew. Chem., Int. Ed. Engl. 1997, 36, 631–634. (c) Gong, L.-Z.; Pu, L.
Tetrahedron Lett. 2001, 42, 7337–7340.
competing hydrolysis of the G3 bromide 7 back to the
corresponding alcohol 16 became a prominent side reaction.
The corresponding G3 iodide was then used as a substitute
for the G3 bromide, but no improvement in reaction yield
was realized. Fortunately, addition of toluene to the reaction
medium gave the di-N-alkylation product Boc-Di-L-G3
(
15) (a) Deb, S. K.; Maddux, T. M.; Yu, L. J. Am. Chem. Soc. 1997, 119,
079–9080. (b) Meier, H.; Lehmann, M. Angew. Chem., Int. Ed. 1998, 37,
ez-Barra, E.; Garcıa-Martınez, J. C.; Merino, S.; del Rey, R.;
guez-L oꢀ pez, J.; S ꢀa nchez-Verd uꢀ , P.; Tejeda, J. J. Org. Chem. 2001, 66,
664–5670. (d) Chow, H.-F.; Ng, M.-K.; Leung, C.-W.; Wang, G.-X. J. Am.
Chem. Soc. 2004, 126, 12907–12915.
16) Xu, Z. F.; Moore, J. S. Angew. Chem., Int. Ed. Engl. 1993, 32, 246–
48.
17) Chow, H.-F.; Ng, K.-F.; Wang, Z.-Y.; Wong, C.-H.; Luk, T.; Lo,
C.-M.; Yang, Y.-Y. Org. Lett. 2006, 8, 471–474.
9
6
43–645. (c) Dı
´
´
´
Rodrı
´
5
(
2
(
(18) Dahlin, N.; Bogevig, A.; Adolfsson, H. Adv. Synth. Catal. 2004, 346,
1101–1105.
J. Org. Chem. Vol. 74, No. 15, 2009 5183

Lo and Chow
JOCArticle
SCHEME 3. Synthesis of Di-L-Gn
SCHEME 4. Synthesis of Mono-L-Gn
and the reaction proceeded extremely slowly, resulting in
a very low product conversion and hence poor product
ield. If a creamy oil-in-water (O/W) emulsion was formed
(Figure 2a), the reaction usually proceeded to give products
in excellent reaction yield. Good results could also be
obtained even though some oil droplets (Figure 2b) or thin
layers of oil (Figure 2c) coexisted with the turbid emulsion.
However, if the reaction mixture became macroscopically
separated phases (Figure 2d,e), the reaction generally failed
to proceed.
Asymmetric Aldol Reactions. The catalytic properties
of the dendritic organocatalysts were examined by con-
ducting the asymmetric aldol reactions using cyclopenta-
none 23 and 3-nitro-benzaldehyde 24 in an emulsion.
The preliminary results of the different dendritic organoca-
talysts were summarized in Table 1. The amphiphilic cata-
1
1
lyst 1 reported by Barbas was also used as a standard for
comparison.
in 56% yield. Subsequent removal of the Boc group by TFA
afforded the desired catalyst Di-L-G3 in 97% yield.
The reductive amination method was used to prepare the
singly dendronized Mono-L-Gn and Mono-S-Gn catalysts
It was found that the various dendritic organocatalysts
were very effective in the aldol reaction. The aldol adduct 25
was formed in 64-93% yields (entries 2-9) that were
comparable to that from using the nondendritic organoca-
talyst 1. Among the three series of compounds, the Di-L-Gn
series (entries 2-4) gave the best product ee values and those
(70-86%) obtained from Di-L-G1 and Di-L-G2 were even
better than that of the organocatalyst 1 (65%). On the other
hand, the Mono-S-Gn series (entries 8-10) gave moderate to
high ee values (52-68%), while the Mono-L-Gn series (en-
tries 5-7) were the least enantioselective catalysts (7-59%).
The results could be rationalized by the higher steric hin-
drance of the Di-L-Gn series, in which there are two bulky
dendron groups to provide the steric shielding. The extre-
mely low ee value (7%) in the case of Mono-L-G1 may be due
to the poor stereodifferentiating effect of the smallest G1
dendron, which is located quite far away from the pyrroli-
dine ring by the longer three-carbon linker. However, as the
generation increases, the size of the dendron becomes larger,
which enables better shielding of the reaction site, and hence,
higher ee values were observed with the Mono-L-G2 and
Mono-L-G3 dendritic catalysts. The Mono-S-Gn series also
offered higher steric hindrance than the Mono-L-Gn series
due to the closer proximity of the dendron to the catalytic
(
Schemes 4 and 5). Oxidation of the alcohols 14-16
with PCC afforded the corresponding aldehydes 17-19 in
8-99% yield, respectively. Treatment of a mixture of the
9
amine 2 and the aldehydes 17-19 in ethanol with NaBH4
gave the monoalkylated products Boc-Mono-L-Gn in
6
5-85% yield, respectively. Subsequent removal of Boc
group afforded the target Mono-L-Gn in 95-97% yield.
The Mono-S-Gn catalysts were also prepared from the
reductive amination of the known shorter chain aldehydes
1
7
2
0-22 and the amine core 2 to give the Boc-protected
monoalkylated products Boc-Mono-S-Gn. Removal of the
Boc protective group by TFA then produced the target
Mono-S-Gn. The overall yields of Mono-S-G1, Mono-S-
G2, and Mono-S-G3 were 67, 61, and 59%, respectively,
from the aldehydes 20-22.
Catalytic Properties. The catalytic properties of the am-
phiphilic dendritic organocatalysts were then examined. It
was found that, to form a stable emulsion, the catalyst and
TFA had to be mixed and sonicated before adding the
substrates. Otherwise, an emulsion might not be formed
5
184 J. Org. Chem. Vol. 74, No. 15, 2009

Lo and Chow
JOCArticle
proline core. However, addition of an extra G1 dendron to
Mono-S-G1 (i.e., Di-S-G1) resulted in a drastic reduction of
product yield (20%), although a relatively good product ee
value (67%) could still be obtained (see Supporting Informa-
tion for details). Hence, increasing steric hindrance although
provided better enantioselectivities, it also retarded the
reaction rate and decreased product yields.
Among these dendritic catalysts, the Di-L-Gn series
were found to form very stable emulsions. Their O/W
emulsions remained stable within 1 h even without stirring,
while those of the Mono-L-Gn separated into two phases
in minutes once stirring was stopped. The Di-L-Gn catalysts
also promoted better product yields and good enantio-
selectivities. Hence, we decided to focus our investigations
mainly on this series of compounds (Table 2). Different
aromatic aldehydes 26 were chosen and reacted with cyclo-
hexanone 27 in creamy O/W emulsion. In all cases, the anti-
aldol 28 was obtained as the major product. For the cases
of electron deficient nitro-substituted benzaldehydes
We also found out that the aldol reaction could still
proceed with a catalytic loading of 1% with no obvious
drop of product yield and ee value (see Supporting Informa-
tion for details). Creamy O/W emulsions could still be
maintained even though the amount of catalyst present
was less.
Relative Catalytic Reactivity. The relative reactivities of
the different generation of dendritic catalysts were deter-
mined to examine the effect of branching on the catalyst
reactivity. The percentage conversions against time were
1
obtained from H NMR measurements. The parameter that
could be extracted from such experiments was the ratio of
aldol product to the remaining aldehyde. The percentage
conversion was calculated by assuming all the aldehyde was
converted to the desired aldol without producing other side
products. However, as the reactions were carried out in an
TABLE 1. Aldol Reactions between Cyclopentanone 23 and 3-Nitro-
benzaldehyde 24 Catalyzed by Various Catalysts
a
(entries 1-8), the reaction proceeded smoothly to give
aldol products in high reaction yields (58-86%). This was
expected as the electron withdrawing nitro group increased
the electrophilicity of the aldehyde. The yield gradually
decreased from the slightly electron deficient 4-bromoben-
zaldehye (entries 9-12) to benzaldehyde (entries 13-16),
and became extremely poor (5-10%) for the electron-rich
4
-methoxybenzaldehyde (entries 17-20). It was noted that
the reaction yields were comparable and, in some cases
entries 2, 3, 10, 15, and 18), even better than those of the
(
nondendritic catalyst 1.
With regard to product ee values, the various dendritic
catalysts behaved differently with different substrates. In
some situations, higher ee values could be achieved (entries 6,
7
, 11, and 18-20) as compared to that of the nondendritic
catalyst 1, while slight inferior ee values were also observed
entries 4, 8, 10, 12, 15, and 16). Some anomalies in the
(
relationship between steric shielding and product ee values
were also observed. For examples, Di-L-G1 was better than
Di-L-G2 for the aldol reaction for 4-methoxybenzaldehyde
(entry 18 vs 19), while Di-L-G2 was better than Di-L-G1
involving 4-bromobenzaldehyde (entry 10 vs 11).
a
Conditions: cyclopentanone 23 (375 μmol), 3-nitrobenzaldehyde 24
(
187 μmol), catalyst (18.7 μmol), and TFA (18.7 μmol) in water (0.5 mL)
b
at 23 °C for 24 h. Isolated yield after column chromatography.
Determined by H NMR analysis of the crude product. Determined
FIGURE 2. (a) Creamy O/W emulsion; (b) O/W emulsion with a
little amount of oil droplets; (c) O/W emulsion with thin layers of oil;
(d) turbid solution that separates into immiscible layers when the
stirring is stopped; (e) insoluble reactants in the reaction mixture.
c
1
d
by chiral-phase HPLC (Chiralcel OD-H) analysis on the major anti-
product.
SCHEME 5. Synthesis of Mono-S-Gn
J. Org. Chem. Vol. 74, No. 15, 2009 5185

Lo and Chow
JOCArticle
TABLE 2. Aldol Reactions between Cyclohexanone 27 and Substituted
Benzaldehydes 26 Catalyzed by Di-L-Gn
effect of the doubly dendronized series but could be ex-
plained by their much better emulsion formation capability
described earlier. Hence, the larger hydrophobic HC den-
dron on one hand retarded the reaction due to larger
steric effect, but on the other hand promoted the formation
of a more stable emulsion and hence an increase of the
reaction rate. The latter effect is similar to a positive dendritic
catalytic effect, that is, increasing dendrimer generation
a
1
9
leads to higher catalyst reactivity. Strictly speaking,
dendritic catalytic effect should be ascribed to the reactivity
pattern of different generations of the same type of com-
pounds, while in our case we are comparing the same
generation of compounds belonging to two different den-
drimer series (i.e., Di-L-Gn vs Mono-L-Gn). However, if
one realizes that the Di-L-Gn series of compounds
actually contain an extra degree of branching at the nitrogen
center, then Di-L-Gn sterically can be considered to be
equivalent to Mono-L-G(nþ1). Hence, in our case we actu-
ally observed reactivity enhancement due to increasing
bulkiness of the HC dendron. The larger HC dendron could
inturnstabilize theresulting emulsion, which was knowntobe
critical in ensuring good product yields and enantioselectiv-
ities. When one examined the reactivity of the three genera-
tions of doubly dendronized catalysts of the same series, it
was shown that the reactivity trend was Di-L-G1>Di-L-G2>
Di-L-G3 (Figure 4). Hence, one could conclude that steric
inhibition was still the dominating factor in determining the
rate of the reaction, while the emulsion stabilization effect
was of second importance. Nonetheless, the product yields
were all >95% for all three generations of catalyst when the
reaction was allowed to extend to 24 h. Hence, the increasing
steric effect just slowed down the reaction rate, but did not
lower the product yield.
b
c
d
entry
catalyst
R
yield (%)
anti/syn
ee (%)
1
2
3
4
5
6
7
1
4-NO
4-NO
4-NO
4-NO
3-NO
3-NO
3-NO
3-NO
4-Br
4-Br
4-Br
4-Br
H
2
2
2
2
2
2
2
2
75
85
80
61
86
85
71
58
50
95
45
38
34
35
45
28
6
83:17
87:13
87:13
74:26
78:22
81:19
88:12
67:33
81:19
70:30
80:20
79:21
78:22
79:21
59:41
61:39
73:27
66:34
64:36
63:37
95
93
93
85
76
91
91
41
66
35
79
13
91
88
60
20
44
80
49
57
Di-L-G1
Di-L-G2
Di-L-G3
1
Di-L-G1
Di-L-G2
Di-L-G3
1
Di-L-G1
Di-L-G2
Di-L-G3
1
Di-L-G1
Di-L-G2
Di-L-G3
1
e
8
9
1
1
1
1
1
1
0
1
2
3
4
5
H
H
H
e
16
17
18
19
20
a
4-OMe
4-OMe
4-OMe
4-OMe
Di-L-G1
Di-L-G2
Di-L-G3
10
5
6
Conditions: cyclohexanone 27 (375 μmol), aldehyde (187 μmol),
catalyst (18.7 μmol), and TFA (18.7 μmol) in water (0.5 mL) at 23 °C
for 24 h. Isolated yield after column chromatography. Determined by
H NMR analysis of the crude product. Determined by chiral-phase
HPLC (Chiralcel OD-H) analysis on the major anti-product. Reaction
time = 72 h.
b
c
1
d
Solvent Effects. We also examined the solvent effects
on the aldol reactions between cyclopentanone 23 and
e
3
-nitrobenzaldehyde 24 using Di-L-G1 (Table 3). A range
of solvents, from nonpolar solvent such as hexane to polar
ones such as water were tested. All reactions proceeded to
give high reaction yields (75-80%) with moderate diaster-
eoselectivities (anti/syn from 2:3 to 2:1). In nonpolar sol-
vents such as hexane (entry 1), where both the ketone
and aldehyde substrates were poorly soluble, low ee (16%)
was observed. For polar nonaqueous solvents such as
ethanol (entry 2) and methanol (entry 3), in which the
reaction was homogeneous, moderate ee values (48%) were
observed. In water (0.5 mL) as an O/W emulsion (entry 4),
a higher ee value (70%) was obtained. Interestingly, decreas-
ing the amount of water to 0.1 mL resulted in the forma-
tion of a water-in-oil (W/O) emulsion (entry 5), which
separated into two layers when stirring was interrupted. In
this case, the product ee value was only 52%. In the absence
of solvent (entry 6), the reaction still proceeded despite of
the high viscosity. However, the product ee value further
dropped to 41%. Hence, a stable O/W emulsion was neces-
sary to produce good product ee values.
emulsion system, the aliquot taken out from the reaction
mixture at each time interval was actually a heterogeneous
mixture containing both organic and aqueous phases.
Hence, there was an uncertainty during samplings, as the
relative volume of aqueous to organic phases could not be
standardized. However, as the aldehyde and the aldol adduct
were present mainly in the organic layer, the variation of the
organic/aqueous volume ratio should not lead to a huge
deviation from the true figures. Hence, the measured relative
reactivity could still provide a semiquantitative picture of the
catalyst efficiency.
The catalytic efficiencies of the different types of G1
and G2 dendritic catalysts were measured (Figure 3). The
nondendritic catalyst 1 was the most effective one as the
reaction was completed within 2 h. In comparison, the
shorter chain catalysts Mono-S-G1 and Di-S-G1 were
clearly inferior as compared to the others. This must be
attributed to a higher steric hindrance of the S series. For
the L series, the order of reactivity exhibited a very inte-
resting trend. It was found that the doubly dendronized
catalysts (L) showed better reactivity than the singly
dendronized catalysts (S) of the same generation (i.e.,
Di-L-G1 > Mono-L-G1 and Di-L-G2 > Mono-L-G2). This
trend was in contradiction to the presumed higher steric
Asymmetric Nitro-Michael Additions. We next exami-
ned the effectiveness of our dendritic organocatalysts in
(
19) (a) Lee, J.-J.; Ford, W. T.; Moore, J. A.; Li, Y. Macromolecules 1994,
27, 4632–4634. (b) Francavilla, C.; Bright, F. V.; Detty, M. R. Org. Lett.
999, 1, 1043–1046. (c) Drake, M. D.; Bright, F. V.; Detty, M. R. J. Am.
1
Chem. Soc. 2003, 125, 12558–12566. (d) Chow, H.-F.; Leung, C.-F.; Wang,
G.-X.; Yang, Y.-Y. C. R. Chimie 2003, 6, 735–745.
5
186 J. Org. Chem. Vol. 74, No. 15, 2009

Lo and Chow
JOCArticle
reusability in another reaction without cross contamination
are very important. Normally, soluble polymer- or dendri-
mer-supported catalysts can be recovered by precipitation
2
0
in a different solvent system or by membrane dialysis if
2
1
they have a relatively high molecular weight. In the case of
our HC organodendritic catalysts, such methods are not
useful as they are oily liquids with a relatively low mole-
cular weight. Hence, a special recovery method was used
in this study. As the HC dendrons are highly nonpolar, they
are highly soluble in nonpolar HC solvents such as hexane
or heptane. Incidentally, the aldol products are more polar
and more soluble in methanol. Making use of the immisci-
bility between nonpolar HC solvent and MeOH, catalyst
recovery was possible by partitioning the reaction mixture
between the two solvent systems.
FIGURE 3. Plot of conversion against time for aldol reactions
between cyclopentanone 23 and 3-nitrobenzaldehyde 24 employing
different dendritic organocatalysts.
The relative amount of our dendritic catalysts in the HC
solvent and MeOH layers after partition (v/v 1:1) could be
1
determined quantitatively by H NMR spectroscopy using
standard addition method (Table 5). As expected, the largest
HC organocatalyst Di-L-G3 possessed the highest partition
ratio in hexane, heptane and octane. On the other hand, the
smallest Di-L-G1 behaved similarly to the nondendritic
catalyst 1, and partitioned mainly in the MeOH layer and
hence could not be extracted and recovered by these non-
polar HC solvents.
For the three nonpolar extraction solvents, heptane showed
better recovery than hexane for Di-L-G2 and Di-L-G3 due to
its higher nonpolar character. Although octane was expected
to give a better recovery due to its even greater nonpolar
character, the recovery was worse than using hexane. One
reason was due to the similar density between octane and
MeOH, which produced problems in the separation of the
two phases. In the cases of hexane and heptane, they have
a relatively larger density difference as compared to MeOH
FIGURE 4. Plot of conversion against time for aldol reactions
between cyclopentanone 23 and 3-nitrobenzaldehyde 24 employing
different Di-L-Gn dendritic organocatalysts.
(Fhexane = 0.659, Fheptane = 0.684, Foctane = 0.708, FMeOH =
.791) and the layers were easily separable. Hence, heptane
0
asymmetric nitro-Michael reactions (Table 4). All reac-
tions proceeded in creamy O/W emulsions and showed good
diastereoselectivities (syn/anti > 82:18). Except Mono-L-G1
and Mono-S-G1 (entries 5, 8, 15, and 18), all other catalysts
gave high reaction yields (61-88%) and good ee values
was chosen as the solvent to recover the Di-L-G2 and Di-L-G3
catalysts after the reactions.
Catalyst recycling by heptane/MeOH partition after
nitro-Michael reaction between cyclohexanone 27 and
trans-β-nitrostyrene 29 (R = H) was then performed. The
catalyst (Di-L-G2 or Di-L-G3) was recovered by extraction
with heptane/MeOH (v/v 3:1). The heptane layer was
then back-washed with MeOH (v/v 1:1), NaOH (1 M) and
saturated NaCl solutions successively and concentrated
in vacuo to obtain the recovered catalyst. It was then
reused in the same nitro-Michael addition (Table 6).
To our delight, the reaction yields (78-87%), product
diastereoselectivities (syn/anti>90:10), and enantioselectiv-
ities (80-84%) could be retained even after five successive
runs. Hence, heptane was found to be a good nonpolar
solvent for extracting the catalysts by partitioning with
MeOH. On the other hand, the nondendritic organocatalyst
(70-90%), and the results were comparable to those using
the nondendritic catalyst 1. Similar to aldol reactions, cata-
lysts Mono-L-G1 and Mono-S-G1 with only one G1 dendron
produced products with lower ee values (31-67%), while
Di-L-G1 gave good results (70-89%). Hence, the extra G1
dendron in Di-L-G1 produced a better stereodifferentiating
and emulsifying effect. It was also found that the electronic
properties of the substituents on the nitrostyrene substrates
did not have any significant effect on the reaction yields.
Hence, both electron donating OMe functionality and
withdrawing NO and Br groups underwent nitro-Michael
2
addition smoothly to afford good product yields (51-88%).
Catalyst Recovery. Because the catalysts were synthe-
sized in multiple steps, the recovery of the catalysts and
1
could not be recycled by heptane extraction as it parti-
tioned in both HC solvent and MeOH layers, the majority
of which was present in the MeOH layer together with the
nitro-Michael adducts. Similarly, catalyst recycling was
also highly efficient for the aldol reactions (see Supporting
Information for details).
(
20) (a) Reetz, M. T.; Lohmer, G.; Schwickardi, R. Angew. Chem., Int.
Ed. Engl. 1997, 36, 1526–1529. (b) Hu, Q.-S.; Pugh, V.; Sabat, M.; Pu, L. J.
Org. Chem. 1999, 64, 7528–7536. (c) Fan, Q.-H.; Chen, Y.-M.; Chen, X.-M.;
Jiang, D.-Z.; Xi, F.; Chan, A. S. C. Chem. Commun. 2000, 789–790. (d)
Maraval, V.; Laurent, R.; Caminade, A.-M.; Majoral, J.-P. Organometallics
2000, 19, 4025–4029. (e) Astruc, D.; Plault, L.; Hauseler, A.; Nlate, S.; Ruiz,
J.; Gatard, S.; Neumann, R. Angew. Chem., Int. Ed. 2004, 43, 2924–2928.
(21) Dijkstra, H. P.; van Klink, G. P. M.; van Koten, G. Acc. Chem. Res.
2002, 35, 798–810.
J. Org. Chem. Vol. 74, No. 15, 2009 5187

Lo and Chow
JOCArticle
TABLE 3. Aldol Reaction between Cyclopentanone 23 and 3-Nitroben-
zaldehyde 24 Catalyzed by Di-L-G1 in Different Solvent Systems
TABLE 5. Partition Ratios ([Catalyst]HC solvent/[Catalyst]MeOH) of
Different Organocatalysts in a Mixture of Nonpolar HC Solvent and
MeOH (v/v 1:1)
a
a
b
yield (%)
c
d
ee (%)
entry
solvents
anti/syn
solvent
1
Di-L-G1
Di-L-G2
Di-L-G3
1
2
3
4
5
6
a
hexane
ethanol
methanol
waterf
79
77
79
80
80
75
40:60
55:45
63:37
69:31
63:37
60:40
16
48
48
70
52
41
hexane
heptane
octane
a
0.1
0.2
0.2
1.2
0.1
0.4
2.1
3.3
1.8
15.7
50.5
6.1
e
water
See Supporting Information for details.
g
Conditions: cyclopentanone 23 (375 μmol), 3-nitrobenzaldehyde 24
a
(
187 μmol), Di-L-G1 (18.7 μmol), and TFA (18.7 μmol) in different
TABLE 6. Asymmetric Nitro-Michael Addition between Cyclohexa-
none 27 and trans-β-Nitrostyrene 29 (R = H) in Water Using Recovered
Organodendritic Catalysts from Heptane/MeOH Partition
b
solvent (0.5 mL) at 23 °C for 24 h. Isolated yield after column
chromatography. Determined by H NMR analysis of the crude
c
1
d
product. Determined by chiral-phase HPLC (Chiralcel OD-H) analysis
b
yield (%)
c
d
ee (%)
catalyst
run
anti/syn
e
on the anti-product. Reaction time was 72 h. 0.1 mL of water used.
f
g
Di-L-G2
Di-L-G2
Di-L-G2
Di-L-G2
Di-L-G2
Di-L-G3
Di-L-G3
Di-L-G3
Di-L-G3
Di-L-G3
a
1
2
3
4
5
1
2
3
4
5
85
84
87
81
83
78
79
79
80
81
6:94
6:94
6:94
8:92
9:91
6:94
6:94
6:94
9:91
8:92
81
80
84
84
82
84
81
83
83
81
Neat without water.
TABLE 4. Nitro-Michael Reactions between Cyclohexanone 27 and
Substituted β-Nitrostyrenes 29 Catalyzed by Various Catalysts
a
Conditions of nitro-Michael addition: cyclohexanone 27 (375 μmol),
trans-β-nitrostyrene 29 (187 μmol), recovered catalyst, and TFA (1 equiv
b
to the catalyst) in water (0.5 mL) at 23 °C for 24 h. Isolated yield after
column chromatography. Determined by H NMR analysis of the
c
1
d
crude product. Determined by chiral-phase HPLC (Chiralcel OD-H)
analysis on the syn-product.
examination of the composition of the heptane layer after the
1
first nitro-Michael addition by H NMR spectroscopy re-
vealed one clear difference between the purity of the recov-
ered Di-L-G2 and Di-L-G3. While the recovered heptane
layer from the Di-L-G3 catalyzed reaction contained only
pure Di-L-G3, the heptane layer from the Di-L-G2 catalyzed
reaction was contaminated with 5 mol % of the nitro-Michael
adduct 30 (R=H) and 1 mol % of trans-β-nitrostyrene. In
principle, the partition coefficients of the nitro-Michael
adduct and trans-β-nitrostyrene should be independent of
the catalyst used. However, due to the much higher lipophi-
licity of Di-L-G3 than Di-L-G2, the former catalyst “satu-
rated” the heptane layer and produced an extrusion effect
that prevented the dissolution of other chemical species in the
same solvent during the extraction process.
b
yield (%)
c
d
ee (%)
entry
catalyst
R
anti/syn
1
2
3
4
5
6
7
8
9
1
H
H
H
H
H
H
H
H
H
H
87
84
80
78
79
88
81
77
83
87
84
61
68
75
59
88
82
51
71
76
82
79
76
66
11:89
10:90
7:93
5:95
7:93
6:94
6:94
14:86
6:94
5:95
5:95
5:95
4:96
4:96
5:95
5:95
5:95
6:94
4:96
5:95
5:95
3:97
18:82
8:92
82
72
83
76
59
83
80
67
86
86
86
71
87
90
31
78
80
41
88
89
73
70
86
89
Di-L-G1
Di-L-G2
Di-L-G3
Mono-L-G1
Mono-L-G2
Mono-L-G3
Mono-S-G1
Mono-S-G2
Mono-S-G3
1
Di-L-G1
Di-L-G2
Di-L-G3
Mono-L-G1
Mono-L-G2
Mono-L-G3
Mono-S-G1
Mono-S-G2
Mono-S-G3
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
a
4-OMe
4-OMe
4-OMe
4-OMe
4-OMe
4-OMe
4-OMe
4-OMe
4-OMe
4-OMe
4-Br
Di-L-G3 was then reused in alternative nitro-Michael
additions with different substrates to see whether any
cross contamination occurred after the recycling pro-
cess (Table 7). In the first, third, and fifth runs, nitro-Michael
addition (reaction A) between cyclohexanone 27 and
trans-β-nitrostyrene was performed, and the substrates
were changed to cyclohexanone 27 and trans-4-methoxy-β-
nitrostyrene 29 (R = 4-OMe) in the second, fourth, and
sixth runs (reaction B). After repeating different nitro-Mi-
chael addition several times, the recovered Di-Li-G3 was
still capable of producing high reaction yields, good product
diastereoselectivities, and enantioselectivities. Most notably,
cross contamination of products from previous runs was
Di-L-G1
1
Di-L-G1
4-Br
2-NO
2-NO
2
2
Conditions: cyclohexanone 27 (375 μmol), nitrostyrene 29
187 μmol), catalyst (18.7 μmol), and TFA (18.7 μmol) in water
(
b
(
0.5 mL) at 23 °C for 24 h. Isolated yield after column chromatography.
c
1
Determined by H NMR analysis of the crude product. Determined
d
by chiral-phase HPLC (Chiralcel OD-H) analysis on the syn-product.
1
not observed as revealed by H NMR analysis of the
crude product mixture.
It is desirable that the recovered catalysts could also be
used in different reactions without cross-contamination. It
is important that the recovered catalyst from the heptane
layer does not contain any starting materials or reaction pro-
ducts carried down from the previous experiment. Interestingly,
Conclusion
Three series of G1-G3 amphiphilic dendritic organo-
catalysts were synthesized and characterized. The structural
5
188 J. Org. Chem. Vol. 74, No. 15, 2009

Lo and Chow
JOCArticle
18-crown-6 and i-Pr NEt (10 equiv) in dry DMF (or DMF/
TABLE 7. Di-L-G3/TFA-Catalyzed Asymmetric Nitro-Michael Addi-
tions with Different Substrates in Water Using Recovered Catalysts
Obtained from Heptane/MeOH Partition
2
a
toluene 1:1) was heated at 60 °C for 24-120 h. The mixture was
2
cooled to 25 °C, diluted with H O and extracted with EtOAc.
The combined extracts were washed with H O, saturated NaCl
2
4
solution, dried (MgSO ), and filtered. After evaporation of
solvent under reduced pressure, the residue was purified by
flash column chromatography on silica gel to afford the target
compound.
Boc-Di-L-G1. Starting from N-Boc-(S)-2-aminomethylpyrro-
2 3
lidine 2 (0.16 g, 0.80 mmol), 5 (0.48 g, 1.73 mmol), K CO (1.09 g,
7
i-Pr NEt (1.38 mL, 7.92 mmol) in dry DMF (5 mL). Upon
.89 mmol), KI (0.29 g, 1.75 mmol), 18-crown-6 (10 mg), and
2
heating for 24 h, the product (0.35 g, 74%) was obtained as
a colorless oil after flash column chromatography (eluent: hex-
ane/EtOAc/Et N 88:11:1). R = 0.23 (hexane/EtOAc 5:1). [R] -
3
f
D
b
yield (%)
c
d
ee (%)
1
run
reaction
anti/syn
61.4 (c = 0.5, CHCl
3
). H NMR (300 MHz, CDCl
3
; rotameric
mixture): δ 0.85 (24 H, d, J = 6.9 Hz), 1.00-1.57 (30 H, m), 1.45
1
2
3
4
5
6
a
A
B
A
B
A
B
81
78
71
82
78
76
17:83
6:94
17:83
7:93
16:84
5:95
88
78
85
81
83
82
(
9 H, s), 1.68-2.08 (4 H, m), 2.08-2.62 (6 H, m), 3.15-3.42 (2 H,
13
m), 3.64-3.95 (1 H, m). C NMR (75 MHz, CDCl ; rotameric
3
mixture): δ 22.6, 22.8, 23.5, 24.7, 28.5, 28.7, 29.3, 31.3, 31.4, 31.5,
36.0, 37.9, 46.3 þ 46.7, 55.6, 56.1, 56.6 þ 57.5, 78.8 þ 79.1, 154.6.
þ
MS (ESI) m/z (%): 594 (100) [(M þ H) ]. HRMS (ESI) calcd for
þ
Conditions of nitro-Michael additions: cyclohexanone 27 (375 μmol),
C H N O , 593.5980; found, 593.5986. Elem anal. Calcd (%)
77
38
2
2
nitrostyrene 29 (187 μmol), recovered catalyst, and TFA (1 equiv to the
catalyst) in water (0.5 mL) for 24 h. Isolated yield after column
76 2 2
for C38H N O : C, 76.96; H, 12.92; N, 4.72. Found: C, 76.83; H,
b
12.95; N, 4.39.
Boc-Di-L-G2. Starting from N-Boc-(S)-2-aminomethylpyrroli-
dine 2 (0.32 g, 1.60 mmol), 6 (1.76 g, 3.32 mmol), K CO (2.19 g,
5.8 mmol), KI (0.58 g, 3.49 mmol), 18-crown-6 (20 mg), and
c
chromatography. Determined by H NMR analysisofthe crude product.
1
d
Determinedbychiral-phase HPLC(ChiralcelOD-H) analysisonthesyn-
product.
2
3
1
2
i-Pr NEt (2.76 mL, 15.8 mmol) in dry DMF (20 mL). Upon
heating for 48 h, the product (1.34 g, 77%) was obtained as a pale
yellow oil after flash column chromatography (eluent: hexane/
effects of the dendron on the emulsion forming abilities,
catalytic efficiencies, product enantioselectivities, and pro-
duct yields were investigated. It was found that increasing
the size and number of dendrons led to higher product
enantioselectivity due to their better steric stereodifferentia-
tion property. The larger hydrophobic HC dendrons
have self-counteracting roles on the catalyst reactivity.
They first retard the reaction due to steric inhibition but
promote the formation of a more stable emulsion and, hence,
result in rate acceleration. Hence, we observed the unusual
catalytic reactivity profile of Di-L-G2 > Mono-L-G2 and
Di-L-G1>Mono-L-G1, that is, increasing branching leads
to higher reactivity. On the other hand, the catalytic reactiv-
ity decreases with increasing generation within the same
series of dendrons, that is, Di-L-G1>Di-L-G2>Di-L-G3.
The optimal catalysts are the G2 dendritic catalysts; most of
them exhibited good reactivities and enantioselectivities
toward asymmetric aldol and nitro-Michael additions in
emulsion systems. In addition, the highest generation G3
dendritic organocatalysts could be recovered by partition
between MeOH and heptane and reused in the same or
different reactions with little loss of reactivity and stereo-
selectivity. Most importantly, no cross contamination was
found when the recovered catalysts were used in different
reactions. Hence, the incorporation of hyperbranched hy-
drophobic dendrons in dendritic organocatalysts was found
to have the advantages of promoting emulsion formation in
water, enhancing the selectivity of the asymmetric reactions,
and facilitating catalyst recovery after the reactions.
EtOAc/Et N 100:5:1). R
3
f
= 0.61 (hexane/EtOAc 5:1). [R]
D
-22.9
). H NMR (300 MHz, CDCl ; rotameric mixture):
1
(
c = 0.6, CHCl
3
3
δ 0.87 (48 H, d, J = 6.6 Hz), 1.00-1.58 (78 H, m), 1.47 (9 H, s),
1
.68-2.08 (4 H, m), 2.08-2.63 (6 H, m), 3.18-3.43 (2 H, m), 3.63-
13
3
3.98 (1 H, m). C NMR (75 MHz, CDCl ; rotameric mixture): δ
22.6, 22.9, 23.8, 24.8, 28.3, 28.5, 28.7, 29.4, 31.4, 31.6, 34.2, 36.0,
37.4, 37.9, 46.3 þ 46.7, 55.7, 56.2, 56.6 þ 57.4, 78.7 þ 78.9, 154.5.
þ
MS (ESI) m/z (%): 1098 (100) [(M þ H) ]. HRMS (ESI) calcd for
þ
C
74
H
149
N
2
O
2
, 1098.1614; found, 1098.1624. Elem anal. Calcd
%) for C H N O : C, 80.95; H, 13.59; N, 2.55. Found: C, 80.69;
(
74 148 2 2
H, 13.68; N, 2.25.
Boc-Di-L-G3. Starting from N-Boc-(S)-2-aminomethylpyrro-
lidine 2 (104 mg, 0.52 mmol), 7 (1.08 g, 1.04 mmol), K CO (0.72
g, 5.21 mmol), KI (0.19 g, 1.14 mmol), 18-crown-6 (7 mg), and i-
Pr NEt (1.0 mL, 5.74 mmol) in dry DMF/toluene (v/v 1:1, 10 mL).
2
3
2
Upon heating for 120 h, the product (0.62 g, 56%) was obtained
as a pale yellow oil after flash column chromatography (eluent:
CH Cl
2
gradient to CH
Cl
-14.8 (c = 0.5, CHCl
; rotameric mixture): δ 0.88 (96 H, d, J =
/EtOAc/Et N 88:11:1). R = 0.54
2
2
2 3 f
1
(
(
hexane/EtOAc 10:1). [R]
D
3
). H NMR
300 MHz, CDCl
3
6
2
.9 Hz), 1.01-1.62 (174 H, m), 1.47 (9 H, s), 1.70-2.10 (4 H, m),
13
C
.10-2.70 (6 H, m), 3.14-3.48 (2 H, m), 3.64-3.97 (1 H, m).
3
NMR (75 MHz, CDCl ; rotameric mixture): δ 23.0, 23.8, 24.1,
2
4.8, 28.4, 28.6, 28.8, 29.5, 31.5, 31.8, 34.3, 34.5, 34.6, 36.1, 37.5,
3
7.8, 38.0, 46.4 þ 46.9, 55.9, 56.3, 56.7 þ 57.4, 78.8 þ 79.1, 154.6.
þ
MS (ESI) m/z (%): 2108 (100) [(M þ H) ]. HRMS (ESI) calcd for
þ
C
H N O
146 293 2 2
, 2107.2882; found, 2107.2864. Elem anal. Calcd
(
%) for C146H292N O : C, 83.19; H, 13.96; N, 1.33. Found: C,
2 2
82.99; H, 14.25; N, 1.17.
General Procedure for the Synthesis of Boc-Mono-L/S-Gn. A
mixture of N-Boc-(S)-2-aminomethylpyrrolidine 2 (1.0 equiv)
and the aldehyde (1.0 equiv) in EtOH was stirred at 25 °C for
Experimental Section
12 h. NaBH (10 equiv) was then added at 0 °C. The mixture was
4
allowed to warm to 25 °C and stirred for 6-24 h. It was acidified
General Procedure for the Synthesis of Boc-Di-L-Gn. A mix-
ture of N-Boc-(S)-2-aminomethylpyrrolidine 2 (1.0 equiv),
the bromide (2.1 equiv), K CO (10 equiv), KI (2.2 equiv), and
to pH = 5 with HCl solution (1 M) and stirred for 15 min. The
pH of the solution was adjusted to 10 by the addition of
saturated KOH solution (2.5 M) at 0 °C. The mixture was then
2
3
J. Org. Chem. Vol. 74, No. 15, 2009 5189

Lo and Chow
JOCArticle
extracted with EtOAc, and the combined extracts were washed
24 h, the product (0.47 g, 62%) was obtained as a pale yellow oil after
flash column chromatography (eluent: hexane/EtOAc 4:1 gradient
with saturated NaCl solution, dried (MgSO ), and filtered. After
4
evaporation of solvent under reduced pressure, the residue was
purified by flash column chromatography on silica gel to afford
the target product.
to EtOAc/Et N 100:1). R
3
f
= 0.29 (hexane/EtOAc 1:1). [R]
D
-19.6
). H NMR (300 MHz, CDCl ; rotameric mixture):
1
(c = 0.5, CHCl
3
3
δ 0.84 (24 H, d, J = 6.6 Hz), 0.96-1.58 (35 H, m), 1.44 (9 H, s),
1.58-2.07 (5 H, m), 2.37-2.90 (4 H, m), 3.08-3.58 (2 H, m), 3.61-
Boc-Mono-L-G1. N-Boc-(S)-2-aminomethylpyrrolidine 2(0.14 g,
.69 mmol), 17 (0.15 g, 0.69 mmol), and NaBH (0.26 g, 6.90 mmol)
1
3
0
4
3
4.08 (1 H, m). C NMR (75 MHz, CDCl ; rotameric mixture): δ
in EtOH (10 mL) were combined. Upon heating for 6 h, the pro-
duct (0.20 g, 73%) was obtained as a pale yellow oil after flash
22.8, 23.8, 23.9, 28.5, 28.6, 29.4, 29.8, 31.3, 31.4, 32.6, 34.2, 36.00,
36.04, 37.9, 38.1, 46.5 þ 46.9, 53.5, 53.9, 57.2, 79.2, 154.8. MS (ESI)
þ
column chromatography (eluent: EtOAc/Et
3
N 100:1). R
f
= 0.24
). H NMR (300 MHz,
; rotameric mixture): δ 0.79 (12 H, d, J = 6.6 Hz), 0.92-1.56
16 H, m), 1.39 (9 H, s), 1.62-2.02 (4 H, m), 2.34-2.88 (4 H, m,
m/z (%): 622 (100) [(M þ H) ]. HRMS (ESI) calcd for
1
þ
(
EtOH). [R]
D
-27.0 (c = 0.25, CHCl
3
C
40
H
81
N
2
O
2
, 621.6293; found, 621.6301. Elem anal. Calcd (%)
CDCl
3
for C40H N O : C, 77.35; H, 12.98; N, 4.51. Found: C, 77.18;
80 2 2
(
H, 13.41; N, 4.21.
Boc-Mono-S-G3.
1
3
3
.10-3.50 (2 H, m), 3.65-3.94 (1 H, m). C NMR (75 MHz,
CDCl ; rotameric mixture): δ 22.7, 23.1, 23.8, 27.4, 28.4, 28.6, 29.3,
9.7, 31.2, 35.9, 37.8, 46.4 þ 46.8, 50.7, 53.3, 57.2, 79.1, 154.89 þ
N-Boc-(S)-2-aminomethylpyrrolidine
2 (94 mg, 0.47 mmol), 22 (0.44 g, 0.47 mmol), and NaBH
4
3
2
1
(0.18 g, 4.76 mmol) in EtOH/toluene (1/2, 30 mL) were
combined. Upon heating for 24 h, the product (0.34 g, 64%)
was obtained as a pale yellow oil after flash column chroma-
tography (eluent: hexane/EtOAc 10:1 gradient to hexane/
þ
54.91. MS (ESI) m/z (%): 397 (100) [(M þ H) ]. HRMS (ESI)
þ
49 2 2
calcd for C24
H N O , 397.3789; found, 397.3795.
Boc-Mono-L-G2. N-Boc-(S)-2-aminomethylpyrrolidine 2(0.14 g,
.71 mmol), 18 (0.33 g, 0.71 mmol), and NaBH (0.27 g, 7.10 mmol)
0
EtOAc/Et N 50:50:1). R
3
f
= 0.20 (hexane/EtOAc 3:1). [R]
D
-
). H NMR (300 MHz, CDCl ; rotameric
4
1
in EtOH (10 mL) were combined. Upon heating for 24 h, the
product (0.30 g, 65%) was obtained as a pale yellow oil after flash
column chromatography (eluent: hexane/EtOAc 10:1 gradient to
EtOAc/Et N 100:1). R = 0.07 (hexane/EtOAc 1:1). [R] -20.3
11.9 (c = 0.5, CHCl
3
3
mixture): δ 0.83 (48 H, d, J = 6.6 Hz), 0.96-1.57 (83 H, m),
1.42 (9 H, s), 1.61-2.00 (5 H, m), 2.29-2.96 (4 H, m), 3.08-
1
3
3
f
D
3
3.48 (2 H, m), 3.59-4.00 (1 H, m). C NMR (75 MHz, CDCl ;
1
(
c = 0.5, CHCl
δ 0.86 (24 H, d, J = 6.6 Hz), 0.98-1.70 (39 H, m), 1.45 (9 H, s),
.70-2.06 (5 H, m), 2.40-2.92 (4 H, m), 3.18-3.50 (2 H, m), 3.70-
3
). H NMR (300 MHz, CDCl
3
; rotameric mixture):
rotameric mixture): δ 22.9, 23.7, 24.0, 28.5, 28.6, 29.3, 29.8,
31.3, 32.8, 34.2, 34.4, 36.0, 37.4, 37.8, 38.4, 46.5 þ 46.9, 53.5,
53.9, 57.3, 79.0, 154.7 þ 154.9. MS (ESI) m/z (%): 1126 (100)
1
3
13
þ
þ
O ,
2
.98 (1 H, m). C NMR (75 MHz, CDCl ; rotameric mixture): δ
[(M
þ
H) ]. HRMS (ESI) calcd for
C
76
H
153
N
2
3
22.8, 23.2, 23.8, 24.0, 27.5, 28.5, 28.7, 29.5, 29.6, 29.9, 31.4, 34.1, 34.2,
1126.1927; found, 1126.1932. Elem anal. Calcd (%) for
76 152 2 2
C H N O : C, 81.06; H, 13.61; N, 2.49. Found: C, 81.14;
3
6.0, 37.4, 37.9, 46.5 þ 46.9, 50.7, 53.4, 57.3, 79.3, 154.9 þ 155.0. MS
þ
(ESI) m/z (%): 650 (100) [(M þ H) ]. HRMS (ESI) calcd for
H, 14.00; N, 2.05.
General Procedure for the Synthesis of Di-L-Gn. A mixture of
TFA (20 equiv) and the Boc-protected amine (1.0 equiv) in
þ
C
42 85 2
H N O
2
, 649.6606; found, 649.6611. Elem anal. Calcd (%)
for C42
3.34; N, 4.24.
Boc-Mono-L-G3. N-Boc-(S)-2-aminomethylpyrrolidine 2 (32 mg,
.16 mmol), 19 (0.16 g, 0.16 mmol), and NaBH (61 mg, 1.61 mmol)
in EtOH/toluene (v/v 1:1, 10 mL) were combined. Upon heating for
4 h, the product (0.15 g, 85%) was obtained as a pale yellow oil after
flash column chromatography (eluent: hexane/EtOAc 3:1 gradient to
hexane/EtOAc/Et N 50:50:1). R = 0.16 (hexane/EtOAc 1:1). [R]
0.7 (c = 0.5, CHCl ). H NMR (300 MHz, CDCl ; rotameric
84 2 2
H N O : C, 77.71; H, 13.04; N, 4.32. Found: C, 77.71; H,
1
CH
was adjusted to 10 by the addition of saturated Na
at 0 °C. The mixture was then extracted with CH
2
Cl
2
was stirred at 25 °C for 12-24 h. The pH of the solution
2
CO
Cl
3
solution
and the
2
0
4
2
combined extracts were washed with saturated NaCl solution,
dried (MgSO ), and filtered. The solvent was evaporated under
reduced pressure to afford the target product.
2
4
-
Di-L-G1. Starting from TFA (2.02 mL, 26.2 mmol) and Boc-
3
f
D
1
1
3
3
Di-L-G1 (0.78 g, 1.32 mmol) in CH
(0.64 g, 99%) was obtained as a yellow oil after stirring for
2 2
Cl (10 mL), the product
mixture): δ0.86 (48 H, d, J=6.6Hz),0.98-1.70 (87 H, m), 1.46 (9 H,
s), 1.72-2.08 (5 H, m), 2.43-2.92 (4 H, m), 3.20-3.53 (2 H, m),
1
12 h. R
= 0.10 (EtOH). [R]
þ12.8 (c = 0.5, CHCl
). H NMR
3
f
D
1
3
3
.64-4.00 (1 H, m). CNMR(75MHz, CDCl
3
;rotamericmixture):
(300 MHz, CDCl ): δ0.87 (24 H, d, J=6.6Hz), 1.00-1.28 (22 H, m),
3
δ 22.9, 23.8, 24.0, 27.5, 28.6, 28.7, 29.5, 29.8, 31.4, 34.2, 34.3, 34.5,
1.28-1.57 (9 H, m), 1.64-1.79 (2 H, m), 1.79-1.94 (1 H, m),
3
6.1, 37.5, 37.7, 37.9, 46.6 þ 46.9, 50.8, 53.4, 57.3, 79.3, 154.9. MS
2.03 (1 H, br s), 2.23-2.50 (6 H, m), 2.76-3.02 (2 H, m), 3.20
þ
13
(ESI) m/z (%): 1154 (100) [(M þ H) ]. HRMS (ESI) Calcd for
(1 H, quintet, J = 6.6 Hz). C NMR (75 MHz, CDCl ): δ 22.8,
3
þ
C
78 157 2
H N O
2
, 1154.2240; found, 1154.2244. Elem anal. Calcd (%)
24.3, 24.9, 28.5, 29.6, 31.3, 31.5, 36.0, 37.9, 45.7, 55.2, 56.4, 59.9.
MS (ESI) m/z (%): 493 (100) [(M þ H) ]. HRMS (ESI) calcd for
þ
for C78
3.73; N, 1.96.
Boc-Mono-S-G1. N-Boc-(S)-2-aminomethylpyrrolidine 2(0.16 g,
.78 mmol), 20 (0.14 g, 0.78 mmol), and NaBH (0.29 g, 7.78 mmol)
in EtOH/toluene (v/v 1:1, 10 mL) were combined. Upon heating for
h, the product (0.20 g, 71%) was obtained as a pale yellow oil after
flash column chromatography (eluent: EtOAc/Et N 100:1). R
). H NMR (300 MHz,
CDCl ; rotameric mixture): δ 0.79 (12 H, d, J = 6.6 Hz), 0.96-1.52
H
156
N
2
O
2
: C, 81.18; H, 13.62; N, 2.43. Found: C, 80.72; H,
þ
1
C
33
H
69
N
2
, 493.5455; found, 493.5457. Elem anal. Calcd (%)
for C33H N : C, 80.41; H, 13.91; N, 5.68. Found: C, 80.21;
68 2
H, 14.09; N, 5.70.
Di-L-G2. Starting from TFA (0.76 mL, 9.86 mmol) and
0
4
6
Boc-Di-L-G2 (1.08 g, 0.98 mmol) in CH
duct (0.97 g, 99%) was obtained as a yellow oil after stirring
for 12 h. R = 0.10 (EtOH). R = 0.12 (EtOH). [R]
þ2.62 (c =
0.5, CHCl ). H NMR (300 MHz, CDCl ): δ 0.86 (48 H, d, J =
2 2
Cl (5 mL), the pro-
3
f
=
1
0
.13 (EtOAc). [R]
D
-38.3 (c = 0.5, CHCl
3
f
f
D
1
3
3
3
(
11 H, m), 1.38 (9 H, s), 1.52-2.02 (4 H, m), 2.18-2.74 (5 H, m),
6.6 Hz), 1.00-1.31 (66 H, m), 1.31-1.57 (13 H, m), 1.64-1.81
13
3
.08-3.42 (2 H, m), 3.60-3.98 (1 H, m). C NMR (75 MHz,
CDCl ; rotameric mixture): δ 22.7, 23.0, 23.8, 28.4, 28.5, 29.6, 35.8,
5.9, 38.1, 38.4, 46.4 þ 47.0, 53.4, 53.6, 57.0, 79.2, 154.7 þ 155.3. MS
(2 H, m), 1.81-1.96 (1 H, m), 2.21-2.50 (6 H, m), 2.50-2.70
1
3
C
(1 H, m), 2.80-3.02 (2 H, m), 3.21 (1 H, quintet, J = 6.6 Hz).
NMR (75 MHz, CDCl ; with overlapping of C signals): δ 22.9,
3
3
3
þ
(
ESI) m/z (%): 369 (100) [(M þ H) ]. HRMS (ESI) calcd for
23.8, 24.3, 24.9, 28.6, 29.7, 31.4, 31.7, 34.3, 36.1, 37.5, 37.9, 45.8,
55.3, 56.5, 59.7. MS (ESI) m/z (%): 998 (100) [(M þ H) ].
þ
þ
C , 369.3476; found, 369.3488. Elem anal. Calcd (%) for
C H N O :C, 71.69;H, 12.03;N, 7.60. Found:C, 71.28;H, 12.39;
22 45 2
H N O
2
þ
HRMS (ESI) calcd for C69
H
141
N
2
, 998.1089; found, 998.1091.
22
44
2 2
N, 7.56.
Boc-Mono-S-G2. N-Boc-(S)-2-aminomethylpyrrolidine 2 (0.25 g,
.25 mmol), 21 (0.54 g, 1.25 mmol), and NaBH (0.47 g, 12.5 mmol)
in EtOH/toluene (v/v 1:1, 10 mL) were combined. Upon heating for
Di-L-G3. Starting from TFA (0.33 mL, 4.28 mmol) and Boc-Di-
L-G3 (0.90 g, 0.43 mmol) in CH Cl (5 mL), the product (0.84 g,
97%) was obtained as a yellow oil after stirring for 24 h. R = 0.12
). H NMR
2
2
1
4
f
1
(hexane/EtOAc 6:1). [R]
D
þ2.3 (c = 0.5, CHCl
3
5
190 J. Org. Chem. Vol. 74, No. 15, 2009

Lo and Chow
JOCArticle
(
300 MHz, CDCl ): δ 0.88 (96 H, d, J=6.6 Hz), 1.00-1.35 (154 H,
NMR (75 MHz, CDCl ): δ 22.8, 23.7, 23.8, 25.2, 28.4, 29.4, 31.2,
3
3
m), 1.35-1.61 (21 H, m), 1.64-1.80 (2 H, m), 1.80-1.99 (1 H, m),
2
31.3, 32.5, 34.1, 35.9, 36.0, 37.8, 37.9, 46.0, 53.9, 54.0, 58.5. MS
(ESI) m/z (%): 521 (100) [(M þ H) ]. HRMS (ESI) calcd for
þ
.16-2.56 (6 H, m), 2.57-2.77 (1 H, br s), 2.80-3.06 (2 H, m), 3.23
13
þ
(1 H, quintet, J = 6.6 Hz). C NMR (75 MHz, CDCl ; with
C H N , 521.5768; found, 521.5763.
35 73
3
2
overlapping of C signals): δ 22.9, 23.8, 24.0, 24.3, 24.9, 28.6, 29.7,
Mono-S-G3. Starting from TFA (0.15 mL, 1.95 mmol) and
Boc-Mono-S-G3 (0.23 g, 0.20 mmol) in CH Cl (2 mL), the
product (0.19 g, 92%) was obtained as a yellow oil after stirring
3
1.4, 31.8, 34.3, 34.4, 34.5, 36.1, 37.5, 37.7, 37.9, 45.8, 55.4, 56.6,
2
2
þ
59.6. MS (ESI) m/z (%): 2008 (100) [(M þ H) ]. HRMS (ESI)
þ
calcd for C1
H N
41 285
, 2007.2357; found, 2007.2366. Elem anal.
for 12 h. R = 0.46 (EtOH/Et N 100:1). [R] þ1.1 (c = 0.5,
2
f
3
D
1
). H NMR (300 MHz, CDCl
Calcd (%) for C141
3.86; H, 14.40; N, 1.25.
Mono-L-G1. Starting from TFA (0.62 mL, 8.07 mmol) and Boc-
Mono-L-G1 (0.32 g, 0.81 mmol) in CH Cl (2 mL), the product
284 2
H N : C, 84.35; H, 14.26; N, 1.40. Found: C,
CHCl
3
3
): δ 0.85 (48 H, d, J =
8
6.6 Hz), 0.98-1.58 (84 H, m), 1.62-1.80 (2 H, m), 1.80-1.96
(1 H, m), 2.38-2.77 (6 H, m), 2.84-3.02 (2 H, m), 3.18-3.36
1
(1 H, m). C NMR (75 MHz, CDCl ; with overlapping of
3
2
2
3
(
R
0.22 g, 95%) was obtained as a yellow oil after stirring for 12 h.
C signals): δ 22.9, 23.8, 24.0, 25.6, 28.5, 29.7, 31.4, 32.9, 34.2,
34.4, 36.0, 37.5, 37.9, 38.3, 46.4, 54.1, 55.0, 58.5. MS (ESI) m/z
1
f
= 0.31 (EtOH/Et
3
N 100:1). [R]
3
D
3
þ9.2 (c = 0.5, CHCl ). H
þ
NMR (300 MHz, CDCl ): δ 0.77 (12 H, d, J = 6.6 Hz), 0.94-1.48
(%): 1026 (100) [(M þ H) ]. HRMS (ESI) calcd for
þ
(
2
16 H, m), 1.48-1.72 (2 H, m), 1.72-1.90 (1 H, m), 1.86 (2 H, br s),
C
71
H
145
N
2
, 1026.1402; found, 1026.1395.
General Procedures of Asymmetric Aldol Reactions. All
13
C
.34-2.58 (4 H, m), 2.72-2.92 (2 H, m), 3.06-3.24 (1 H, m).
; with overlapping of C signals): δ 22.85,
2.87, 25.8, 27.4, 28.6, 29.9, 31.4, 36.1, 38.0, 46.6, 50.9, 55.1, 58.4.
NMR (75 MHz, CDCl
2
3
aldol reactions were carried out under air in a closed system.
TFA (1.4 μL, 18.7 μmol) was added to a mixture of the catalyst
þ
MS (ESI) m/z (%): 297 (100) [(M þ H) ]. HRMS (ESI) calcd for
(18.7 μmol) in H O (0.5 mL) at 23 °C with sonication for 15 min.
2
þ
C
19 41
H N
2
, 297.3264; found, 297.3270.
Ketone (375 μmol) and aldehyde (187 μmol) were then sequen-
Mono-L-G2. Starting from TFA (0.35 mL, 4.58 mmol) and Boc-
Mono-L-G2 (0.30 g, 0.46 mmol) in CH Cl (2 mL), the product
tially added. The mixture was sonicated and then stirred vigor-
ously in a closed vial for 24 h to maintain an emulsion. The
reaction was quenched by adding EtOAc/saturated NaCl solu-
tion (v/v 1:1, 4 mL) to allow for phase separation. The organic
layer was collected and the aqueous layer was extracted with
EtOAc (2 ꢀ 1 mL). The combined organic extracts were washed
2
2
(
R
0.24 g, 95%) was obtained as a yellow oil after stirring for 12 h.
1
f
= 0.31 (EtOH/Et
3
N 100:1). [R]
3
D
3
þ5.7 (c = 0.5, CHCl ). H
NMR (300 MHz, CDCl ): δ 0.83 (24 H, d, J = 6.6 Hz), 0.96-1.57
(
2
40 H, m), 1.57-1.79 (2 H, m), 1.79-1.86 (1 H, m), 2.42-2.70 (4 H, m),
1
3
.80-2.96 (2 H, m), 3.01 (2 H, br s), 3.18-3.36 (1 H, m). C NMR
): δ 22.8, 23.7, 25.6, 27.2, 28.5, 29.7, 31.3, 31.4,
4.06, 34.14, 36.0, 37.4, 37.8, 46.3, 50.8, 54.6, 58.4. MS (ESI) m/z
with saturated NaCl solution, dried (MgSO ), and filtered. After
4
(75 MHz, CDCl
3
evaporation of solvent under reduced pressure, the diastereos-
electivity (anti/syn) of the crude product was determined by H
1
3
þ
þ
,
(
%): 549 (100) [(M þ H) ]. HRMS (ESI) calcd for C37
49.6081; found, 549.6080.
Mono-L-G3. Starting from TFA (30 μL, 0.38 mmol) and Boc-
Mono-L-G3 (44 mg, 38 μmol) in CH Cl (2 mL), the product
H N
77 2
NMR analysis. The sample was purified by passing through a
short column of silica gel to remove the faster running and
baseline materials to obtain the product yield. The ee value of
the anti-adduct was determined by chiral HPLC (Chiralcel
OD-H) analysis.
5
2
2
(
R
39 mg, 97%) was obtained as a yellow oil after stirring for 12 h.
1
f
= 0.12 (EtOH/Et
3
N 100:1). [R]
3
D
3
þ2.6 (c = 0.5, CHCl ). H
General Procedures of Asymmetric Nitro-Michael Additions.
All nitro-Michael additions were carried out under air in
a closed system. TFA (1.4 μL, 18.7 μmol) was added to
NMR (300 MHz, CDCl ): δ 0.85 (48 H, d, J = 6.6 Hz), 0.97-1.60
(
2
88 H, m), 1.60-1.81 (2 H, m), 1.81-1.88 (1 H, m), 2.48 (2 H, br s),
13
C
.50-2.70 (4 H, m), 2.84-2.98 (2 H, m), 3.18-3.34 (1 H, m).
; with overlapping of C signals): δ 22.9,
3.8, 23.9, 25.8, 27.3, 28.5, 29.8, 31.4, 34.2, 34.3, 34.4, 36.0, 37.5,
7.6, 37.9, 46.5, 50.9, 55.0, 58.4. MS (ESI) m/z (%): 1054 (100)
a mixture of catalyst (18.7 μmol) in H
with sonication for 15 min. Ketone (375 μmol) and nitrostyrene
187 μmol) were then sequentially added. The mixture
2
O (0.5 mL) at 23 °C
NMR (75 MHz, CDCl
2
3
3
(
was sonicated and then stirred vigorously in a closed vial
for 24 h to maintain an emulsion. The reaction was quenched
by adding EtOAc/saturated NaCl solution (v/v 1:1, 4 mL)
to allow for phase separation. The organic layer was collected
and the aqueous layer was extracted with EtOAc (2 ꢀ 1 mL).
The combined organic extracts were washed with saturated
þ
þ
, 1054.1715;
[
(M þ H) ]. HRMS (ESI) calcd for C73
found, 1054.1723.
Mono-S-G1. Starting from TFA (0.19 mL, 2.50 mmol) and Boc-
Mono-S-G1 (92 mg, 0.25 mmol) in CH Cl (2 mL), the product
63 mg, 94%) was obtained as a yellow oil after stirring for 12 h.
H N
149 2
2
2
(
1
R = 0.37 (EtOH/Et N 100:1). [R] þ8.8 (c = 0.5, CHCl ). H
f
3
D
3
NaCl solution, dried (MgSO ), and filtered. After evapora-
4
NMR (300 MHz, CDCl ): δ 0.71 (12 H, d, J = 6.9 Hz), 0.86-1.42
3
tion of solvent under reduced pressure, the diastereoselectivity
(
1
syn/anti) of the crude product was determined by H NMR
(
2
12 H, m), 1.50-1.71 (2 H, m), 1.71-1.87 (1 H, m), 2.24-2.40 (2 H, m),
.44 (1 H, dd, J = 12.0 and 4.8 Hz), 2.54 (1 H, dd, J = 12.0 and
.8 Hz), 2.75-2.95 (2 H, m), 3.10-3.28 (1 H, m), 3.45 (2 H, br s).
analysis. The sample was purified by passing through a short
column of silica gel to remove the faster running and baseline
materials to obtain the product yield. The ee value of the syn-
product was determined by chiral HPLC (Chiralcel OD-H)
analysis.
4
1
3
3
C NMR (75 MHz, CDCl ): δ 22.5, 25.1, 28.2, 29.2, 29.5, 35.70,
3
5.74, 38.2, 45.9, 53.7, 53.9, 58.3. MS (ESI) m/z (%): 269 (100)
þ
þ
, 269.2951; found,
[
2
(M þ H) ]. HRMS (ESI) calcd for C17
69.2960.
Mono-S-G2. Starting from TFA (0.24 mL, 3.06 mmol) and
Boc-Mono-S-G2 (0.19 g, 0.31 mmol) in CH Cl (2 mL), the product
0.15 g, 98%) was obtained as a yellow oil after stirring for 12 h.
H N
37 2
Acknowledgment. We thank the RAC of The Chinese
University of Hong Kong for the financial support.
2
2
(
1
R = 0.40 (EtOH/Et N 100:1). [R] þ2.0 (c = 0.5, CHCl ). H
f
3
D
3
1
13
NMR (300 MHz, CDCl
3
): δ 0.80 (24 H, d, J=6.6 Hz), 0.96-1.54
Supporting Information Available: H and C NMR spec-
tra and SEC chromatograms of all new compounds. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(
2
36 H, m), 1.61-1.81 (2 H, m), 1.81-1.98 (1 H, m), 2.34-2.52 (2 H, m),
.54 (1 H, dd, J = 11.7 and 8.1 Hz), 2.64 (1 H, dd, J = 11.7 and 4.5
13
C
Hz), 2.84-3.06 (2 H, m), 3.20-3.43 (1 H, m), 3.82 (2 H, br s).
J. Org. Chem. Vol. 74, No. 15, 2009 5191