Diastereoselective addition of organolithiums to 1,3-oxazolidines complexed with aluminum tris(2,6-diphenylphenoxide) (ATPH)

Article abstract of DOI:10.1016/j.tet.2004.12.036

1,3-Oxazolidines were easily obtained by condensation of N-substituted (R)-phenylglycinol with aldehydes. Addition of organolithium reagents to 1,3-oxazolidines by complexation with the bulky Lewis acid aluminum tris(2,6-diphenylphenoxide) (ATPH) readily produced the corresponding chiral amines with good yield and high diastereoselectivity. The configuration of the new stereogenic center was shown to be opposite to that of adducts obtained for the same 1,3-oxazolidines using Grignard reagents. The best diastereoselectivity was achieved using N-isopropyl-1,3-oxazolidines. The mechanism of addition was deduced by determining the stereochemistry of the iminium-aluminum complex by NOE experiments.

Full text of DOI:10.1016/j.tet.2004.12.036

Tetrahedron 61 (2005) 1731–1736
Diastereoselective addition of organolithiums to 1,3-oxazolidines
complexed with aluminum tris(2,6-diphenylphenoxide) (ATPH)
*
Takayasu Yamauchi, Hiroyuki Sazanami, Yuuichi Sasaki and Kimio Higashiyama
Institute of Medicinal Chemistry, Hoshi University, Ebara 2-4-41, Shinagawa, Tokyo 142-8501, Japan
Received 12 October 2004; revised 14 December 2004; accepted 14 December 2004
Available online 11 January 2005
Abstract—1,3-Oxazolidines were easily obtained by condensation of N-substituted (R)-phenylglycinol with aldehydes. Addition of
organolithium reagents to 1,3-oxazolidines by complexation with the bulky Lewis acid aluminum tris(2,6-diphenylphenoxide) (ATPH)
readily produced the corresponding chiral amines with good yield and high diastereoselectivity. The configuration of the new stereogenic
center was shown to be opposite to that of adducts obtained for the same 1,3-oxazolidines using Grignard reagents. The best
diastereoselectivity was achieved using N-isopropyl-1,3-oxazolidines. The mechanism of addition was deduced by determining the
stereochemistry of the iminium–aluminum complex by NOE experiments.
q 2004 Elsevier Ltd. All rights reserved.
1. Introduction
addition of Grignard reagents (Scheme 1). Other groups
have also reported that some reactions with ATPH resulted
in the reversal events of diastereoselectivity.4
The diastereoselective addition of organometallic reagents
to the C]N bond of chiral imines and their derivatives is
useful for the asymmetric synthesis of chiral amines.1 We
have previously described a synthetic method for stereo-
selective preparation of both enantiomers of chiral amines
from a single-enantiomer source, (R)-phenylglycinol, pro-
ceeding via the diastereoselective addition of Grignard
reagents to 1,3-oxazolidines with excellent yield and
diastereoselectivity.2 It was previously alleged that addition
of Grignard reagents occurred after formation of the ring-
opened iminium intermediate, but addition of an organo-
lithium reagent to 1,3-oxazolidine did not proceed for the
unopened ring. It was considered that the reaction required
activation to open the 1,3-oxazolidine ring. We tried to react
1,3-oxazolidines with organolithium reagents using various
Lewis acids. Aluminum compounds might be effective
additives to facilitate the reaction. One additive, bulky C3
symmetrical ATPH, has been shown to have unique
properties in various reactions by Yamamoto.3 ATPH has
a small opening in the ligand sphere and is known to give
stable complexes with carbonyl compounds. Herein we
report the diastereoselective addition of organolithium to
1,3-oxazolidine via activation with ATPH. Interestingly, the
absolute configuration of the adducts obtained in the
presence of ATPH was the opposite to that obtained by
2. Results and discussion
2.1. Addition of MeLi to 1a using various Lewis acids
For the addition of MeLi to 1,3-oxazolidines, an activator
such as a Lewis acid is needed for cleavage of the 1,3-
oxazolidine ring. To activate a diastereomer mixture of 1,3-
oxazolidine 1a,2a prepared easily from (R)-phenylglycinol,
we tried various Lewis acids as additives (Scheme 2,
Table 1). As expected, addition of MeLi to 1a did not
proceed without a Lewis acid (run 1). Some Lewis acids
provided methylation to the 1,3-oxazolidine but with low
yields and diastereoselectivity (runs 2, 3, and 8). Dia-
stereoselective addition of MeLi was possible in the
presence of MgBr2 and Me3Al (runs 9 and 12). Interest-
ingly, the major adduct of methyl addition using ATPH,
(R,R)-2a, differed from that obtained using the other Lewis
acids (runs 13–16). This result also differed from previous
research in which addition of MeMgBr to 1a gave (S,R) 2a
in 94% yield and 68% de.2a It was assumed that the change
in diastereoselectivity was caused by a virtually blocking of
the reaction site due to the bulky structure of ATPH. After a
series of activating experiments, the optimum activation
time of 1a with ATPH was found to be 2 h at rt. When the
reaction time was prolonged, it resulted in a decreased yield
(runs 13–15). As ATPH showed encouraging activity,
Keywords: Lewis acid; Phenylglycinol; NOE experiment; Iminium–
aluminum complex; Allylic strain.
* Corresponding author. Tel./fax: C81 3 5498 5768;
e-mail: yamauchi@hoshi.ac.jp
0040–4020/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2004.12.036

1732
T. Yamauchi et al. / Tetrahedron 61 (2005) 1731–1736
Scheme 1.
Scheme 2.
Table 1. Addition of MeLi to 1a with Lewis acid
Run
Lewis
acid
Activation
time (h)
Reaction
temperature (8C)
Reaction
time (h)
Yield
(%)
Ratio
(R,R/S,R)a
1
—
—
1
1
2
2
2
2
2
2
2
1
2
1
1
1
2
rt
rt
rt
20
20
20
20
20
20
20
20
20
20
20
20
20
72
168
20
NR
34
17
NR
NR
NR
NR
41
—
2b
3b
4
BF3OEt2
BCl3
SnCl2
MnBr2
Et2Zn
Ln(Otf)3
Yb(Otf)3
MgBr2
YCl3
Me3Al
Me3Al
ATPH
ATPH
ATPH
ATPH
43:57
38:62
—
—
—
K50
K50
K50
K50
K50
K50
K50
rt
K50
K50
K50
K50
K50
5
6
7
8
—
37:63
15:85
—
9
68
10
11b
12b
13
14
15
16
NR
Trace
72
62
56
—
20:80
79:21
67:33
80:20
78:22
19
86
a Estimated by 1H NMR spectrum.
b This reaction was carried in THF solvent.
further research looked into the effect of the N-substituent of
1,3-oxazolidines at K50 8C.
of 1a–h were confirmed to be inseparable mixtures in
thermodynamic equilibrium differing at the 2 position of the
1,3-oxazolidine ring, and their ratios in CDCl3 were
1
determined from the H NMR peak intensities of the 2-H
2.2. Diastereoselective additions of organolithium
reagents to N-substituted 1,3-oxazolidines complexed
with ATPH
of 1,3-oxazolidine. Addition of organolithium reagents to
1a–h with ATPH as the Lewis acid gave 2a–e in 62–98%
yield with 78:22wO99:1 diastereoselectivity. The adducts
obtained with ATPH, with the exception of substrate 1b,
showed opposite diastereoselectivities to the adducts
obtained with Grignard reactions. The isomer ratios of the
To probe the influence of the N-substituent of 1,3-
oxazolidine, organolithium reagents were added to various
1,3-oxazolidines complexed with ATPH (Scheme 3,
Table 2). N-Substituted 1,3-oxazolidines (1a–c,2a 1d,2b,c
1f–h2a) were prepared from (R)-phenylglycinol in three
steps as noted in the literature. 1e was also prepared from
(R)-phenylglycinol in the same manner. The diastereomers
1
adducts were determined from the H NMR peak intensity
of the 2-Me. The absolute configurations of 2a–c,2a 2d2b,c
were previously reported. Treatment of the single isomers
(R,R)-2e and (S,R)-2e with TFA gave (R,R)-3 in 77% yield

T. Yamauchi et al. / Tetrahedron 61 (2005) 1731–1736
1733
Scheme 3.
Table 2. Addition of R3Li to 1a–h with ATPH
(cf. Grignard reaction)a
Run
Substrate
R1
R2
R3
Yield
(%)
Ratio
(R,R/S,R)b
|
|
|
|
|
|
|
|
|
|
Ratio
(R,R/S,R)b
1
2
3
4
5
6
7
8
1a (88:12)
1b (97:3)
1c (95:5)
1d (89:11)
1e (90:10)
1f (90:10)
1g (98:2)
1h (98:2)
Bn
Me
Ph
Ph
Ph
Ph
Ph
Me
Me
Me
Me
Me
Me
Me
Me
Ph
86
98
86
62
94
90
83
80
78:22
19:81
97:3
84:16
97:3
3:97
16:84c
34:66c
3:97c
i-Pr
Diphenylmethyl
2,4,6-Trimethylbenzyl
Bn
Me
i-Pr
6:94d
6:94
77:23c
78:22c
88:11c
Ph
Ph
27:73
1:O99
a Reaction with R3MgBr in THF.
b Estimated by 1H NMR spectrum.
c See Ref. 2a.
d See Ref. 2c.
and (S,R)-3 in 84% yield, respectively. The stereochemistry
of 3 was established by comparing H NMR spectra with
aluminum complex (4) at the 1 position is not certain, and
the diastereoselective process as a whole has not yet been
elucidated. To examine the mechanism of addition we
determined the geometry of the iminium–aluminum
complex by NOE experiment.
1
published data2a–d (Scheme 4).
A model compound, N-isopropyl-2-methyl-1,3-oxazolidine
(1h), was preferred to the comparative intelligible chart.
The H NMR spectra of iminium–aluminum complex (4h)
1
prepared from 1h under usual conditions was assigned by
comparison of the decoupling spectra (Fig. 1) with an
equivalent C20 deuterated compound additionally pre-
pared.6 Trace a shows the high-field region of the spectra
of non-deuterated 4h. Traces b–d show the spin decoupling
spectrum acquired by irradiating H20a, H20b, and H100,
respectively. In traces b and c, irradiation at H20a and H20b
was reflected by a change in the H10 peak from a doublet–
doublet to a doublet. Trace d shows that irradiation at H100
converted the doublet at H200 into a singlet. Based on these
experiments, assignment of the 1H NMR spectra of 4h was
judged to be consistent. Both NOE difference experiments
identified correlation between H2 and H10 at rt (Fig. 2). The
iminium–aluminum complex with an N-isopropyl group
(4h) was found to adopt the Z form in CDCl3 (Fig. 3). This
result was unexpected because the geometry seemed to be
more unstable, as the steric repulsion of the phenethyl group
with ATPH would be expected to be greater than that with
the isopropyl group. The stereochemistry at the 2 and 3
Scheme 4.
2.3. Discussion about the diastereoselective addition of
organolithium reagents to 1,3-oxazolidines with ATPH
based on the geometry of the iminium–aluminum
complex
The mechanism of the ring opening of chiral 1,3-
oxazolidines has been previously described.5 First, the
metal coordinates to oxygen, and then the C–O bond of the
oxazolidine ring is cleaved. As a result, an iminium–metal
complex intermediate is formed, with addition of the
organometallic reagent giving the chiral amine. We
considered the mechanism of addition of organolithium
reagents to 1,3-oxazolidines with ATPH to be similar
(Scheme 5). However, the geometry of the iminium–
Scheme 5.

1734
T. Yamauchi et al. / Tetrahedron 61 (2005) 1731–1736
Figure 2. Trace a shows the partial one-dimensional 270 MHz 1H NMR
spectrum of 4h in CDCl3 at 22 8C. Traces b and c are the steady-state NOE
difference spectra obtained, when H2 and H10 are saturated.
Figure 3.
Figure 1. Partial 270 MHz 1H NMR spectrum of 4h in CDCl3 at 22 8C.
(a) Original. (b) Decoupling at 1.6 ppm. (c) Decoupling at 2.7 ppm.
(d) Decoupling at 3.0 ppm.
positions of 1,3-oxazolidine might have been important in
setting the geometry of the iminium–metal complex.
Further, due to the effect of the 1,3-allylic strain, the
conformation at C10 of 4h is fixed. As a result, the bulky
ATPH would situate on the si face and the organolithium
reagent would attack the iminium–aluminum complex from
the re face, avoiding ATPH to give (S,R)-2c (Scheme 6). We
are still unsure of the exact reason why only ATPH
produced this effect when other bulky Lewis acids did not. It
may be a result of the remarkable properties of the C3
symmetrical ATPH, an aluminum center surrounded by
bulky ligands in which the aluminum ‘peeks out’ from a
small opening in the ligand sphere. Supposing the
mechanism by the observations, the geometry of 4h related
with the configuration at 2 position of oxazolidine (1h) and
diastereoselectivity of the addition could be a clear
explanation. However, in the reactions of 1a-1g, the cause
Scheme 6.

T. Yamauchi et al. / Tetrahedron 61 (2005) 1731–1736
1735
of diastereoselectivity is imprecise, because we were unable
to characterize their complexes.
(MCKCH2OH), 133 (base peak). IR (CHCl3, cmK1): 3440
(O–H, N–H). 1H NMR (CDCl3) d: 1.42 (1H, br), 2.25 (9H,
s), 2.69 (1H, br), 3.47–3.76 (4H, m), 3.83 (1H, dd, JZ9.2,
4.8 Hz), 6.84 (2H, s), 7.29–7.42 (5H, m). 13C NMR (CDCl3)
d: 19.39q, 20.95q, 45.53t, 65.40d, 66.56t, 127.30d, 127.66d,
128.62d, 129.06d, 133.33s, 136.59s, 137.00s, 140.71s. Anal.
Calcd for C18H23NO: C, 80.26; H, 8.61; N, 5.20. Found: C,
80.01; H, 8.95; N, 4.96. A mixture of above compound
(5.38 g, 20 mmol) and benzaldehyde (6.36 g, 60 mmol) in
benzene (100 mL) was refluxed for 20 h with a Dean–Stark
trap. After being cooled, the mixture was concentrated
under reduced pressure. The residue was distilled (231 8C,
4 mm Hg) to afford 1e (88%, 90:10 mixture) as colorless oil.
[a]2D0ZK7.40 (c 1.00, CHCl3). MS m/z: CI, 358 (MCC1,
basepeak);EI, 357(MC), 133(basepeak). IR(CHCl3, cmK1):
3030, 2950, 2860 (C–H). 1H NMR (CDCl3) d: major
component; 2.03 (3H, s), 2.09 (6H, s), 3.64 (1H, d, JZ
12.4 Hz), 3.68 (1H, d, JZ12.4 Hz), 3.97 (2H, m), 4.32 (1H,
m), 5.12 (1H, s), 6.41 (2H, s), 7.12–7.41 (10H, m). 13C NMR
(CDCl3) d: major component; 20.33q, 20.66q, 49.59t, 69.05d,
74.52t, 98.66d, 127.15d, 127.35d, 127.73d, 127.77d, 127.88d,
128.44d, 128.48d, 130.77s, 136.21s, 137.20s, 139.98s,
140.11s. HRMS calcd for C25H27NO: 357.2093. Found:
357.2071.
3. Conclusion
1,3-Oxazolidines were reacted with organolithium reagents
using the bulky Lewis acid ATPH. The reactions were
achieved with high yield and high diastereoselectivity, and
the products showed opposite diastereoselectivity to
products of Grignard reaction. The best diastereo-
selectivities were obtained for addition of 1,3-oxazolidines
having N-isopropyl and 2,4,6-trimethylbenzyl groups.
Chiral amines could be synthesized with opposite dia-
stereoselectivity from a chiral 1,3-oxazolidine depending on
whether Grignard reagents or ATPH-organolithium
reagents were used. ATPH was shown to have activating
ability due to effective coordination of the iminium–
aluminum complex with the N,O-acetal. A variation of
this method may be useful for the asymmetric synthesis of
compounds with medical applications, including physio-
logically active natural products.
4. Experimental
4.1. General
4.2. General procedure for the addition of organolithium
reagent to 1a–h with ATPH
A mixture of 2,6-diphenylphenol (0.55 g, 2.25 mmol) and
Me3Al (1.75 mL, 0.75 mmol; 1 M in hexane) in dry CH2Cl2
(2 mL) was stirred at rt under nitrogen for 30 min to afford
the solution of ATPH.6 To this solution was added the
solution of oxazolidine (1a–h) (0.5 mmol) in dry CH2Cl2
(3 mL) and stirred at rt for 2 h. The solution was cooled to
K50 8C, and organolithium (1.5 mL, 1.5 mmol, 1 M
solution) was added dropwise to it. After being stirred at
K50 8C for 20 h, the reaction mixture was treated with a
small amount of water, and the resulting white precipitate
was filtered off. The filtrate was dried over anhydrous
Na2SO4 and concentrated under reduced pressure. The
residue was subjected to column chromatography on silica
gel with hexane–ether (2:1) to give a diastereomeric mixture
of amine (2a–e).
Melting points were measured with a Yanagimoto Micro
melting Point apparatus without collection. IR spectra were
recorded on a 215 Hitachi Granting IR spectrophotometer.
1H and 13C NMR spectra were obtained on a JEOL GSX 270
instrument, and chemical sifts are reported in ppm on the
d-scale from internal Me4Si. MS spectra were measured
with a JEOL JMS D-300 spectrometer by using the chemical
ionization (CI) with isobutene and the electron impact (EI)
methods. Elemental analyses were performed on a Perkin–
Elmer 240-B instrument. Optical rotation were taken with a
JASCO-DIP-370 polarimeter at rt. Sibata Glass Tube Oven
GTO-350RD was used as distillation apparatus. Column
chromatography was performed on silica gel (45–75 mm,
Wakogel C-300). The reaction solvents were prepared as the
following. THF was distilled over potassium metal.
Dichloromethane was distilled over phosphorus pentoxide.
Ether and toluene were distilled over sodium metal.
4.2.1. (1R,10R)-N-20-Hydroxy-10-phenylethyl-N-2,4,6-tri-
methylbenzyl-1-phenylethylamine (R,R-2e). Yield 86%
(97:3 mixture). Diastereomers were separated by column
chromatography on silica gel with hexane–ether (3:1) to
give pure (R,R)-2e as colorless oil. [a]2D2ZK115.5 (c 1.27,
CHCl3). MS m/z: CI, 374 (MCC1, base peak); EI, 373
(MC), 342 (MCKCH2OH), 133 (base peak). IR (CHCl3,
cmK1): 3500 (O–H). 1H NMR (CDCl3) d: 1.28 (3H, d, JZ
7.1 Hz), 1.82 (1H, br), 2.24 (3H, s), 2.28 (6H, s), 3.39 (1H,
m), 3.85 (4H, m), 4.01 (1H, q, JZ7.1 Hz), 6.84 (2H, s),
7.20–7.46 (10H, m). 13C NMR (CDCl3) d: 13.64q, 20.15q,
20.82q, 45.44t, 54.96d, 62.39t, 62.94d, 126.95d, 127.60d,
128.34d, 128.36d, 2!129.54d, 132.13s, 136.62s, 138.03s,
139.06s, 144.42s. Anal. Calcd for C26H31NO: C, 83.60; H,
8.37; N, 3.75. Found: C, 83.64; H, 8.32; N, 3.64.
4.1.1. (2R,4R)-N-(2,4,6-Trimethylbenzyl)-2,4-diphenyl-
1,3-oxazolidine (1e). A mixture of (R)-phenylglycinol
(6.85 g, 50 mmol) and 2,4,6-trimethylbenzylaldehyde
(7.41 g, 50 mmol) in benzene (100 mL) was refluxed for
1 h with a Dean–Stark trap. After being cooled, the mixture
was concentrated under reduced pressure, and the residue
was dissolved in methanol (100 mL). To this solution was
added portionwise NaBH4 (4.73 g, 125 mmol) at rt. After
the reaction mixture was stirred for 40 min, it was added
with water (100 mL) and the aqueous layer was extracted
with CH2Cl2 (3!50 mL). The combined extracts were
washed with brine, dried over anhydrous Na2SO4, and
concentrated under reduced pressure. The residue was
recrystallized (EtOAc–hexane) to afford (R)-N-2,4,6-tri-
methylbenzylphenylglycinol (89%) as colorless needles.
Mp 98–99 8C. [a]2D2ZK72.5 (c 1.00, CHCl3). MS m/z:
CI, 270 (MCC1, base peak); EI, 269 (MC), 238
4.2.2. (1S,10R)-N-20-Hydroxy-10-phenylethyl-NK2,4,6-
trimethylbenzyl-1-phenylethylamine (S,R-2e). Methyl-
magnesium bromide (0.5 mL, 1.5 mmol, 3 M in ether) was

1736
T. Yamauchi et al. / Tetrahedron 61 (2005) 1731–1736
added dropwise to a stirred solution of oxazolidine (1e)
(0.5 mmol) in dry THF (5 mL) at rt under nitrogen over
10 min period. After the reaction mixture was stirred for
20 h, it was quenched with a small amount of water and
diluted with ether (10 mL). The resulting white precipitate
was filtered off, and the filtrate was washed with saturated
aqueous NH4Cl (10 mL). The aqueous phase was extracted
with ether (2!10 mL). The combined organic extract was
washed with brine, dried over anhydrous Na2SO4, and
concentrated under reduced pressure to leave a oily residue,
which was subjected to column chromatography on silica
gel with hexane–ether (2:1) to give a diastereomeric mixture
of amine (2e) (85% yield, 94:6 mixture). Diastereomers
were separated by column chromatography on silica gel
with hexane–ether (3:1) to give pure (S,R)-2e as
colorless oil. [a]2D4ZK68.0 (c 1.40, CHCl3). MS m/z: CI,
374 (MCC1, base peak); EI, 373 (MC), 342 (MC
KCH2OH), 133 (base peak). IR (CHCl3, cmK1): 3440 (O–H).
1H NMR (CDCl3) d: 1.51 (3H, d, JZ7.1 Hz), 1.55 (1H, br),
2.05 (6H, s), 2.22 (3H, s), 3.66 (1H, d, JZ12.7 Hz), 3.86
(1H, dd, JZ9.4, 4.9 Hz), 4.01 (1H, d, JZ12.7 Hz), 4.02
(1H, q, JZ7.1 Hz), 4.15 (1H, dd, JZ10.7, 4.9 Hz), 4.26
(1H, dd, JZ10.7, 9.4 Hz), 6.77 (2H, s), 6.87–7.01 (4H, m),
7.13–7.29 (6H, m). 13C NMR (CDCl3) d: 16.03q, 19.75q,
20.78q, 43.82t, 54.71d, 61.10t, 61.65d, 126.49d, 126.99d,
127.57d, 128.08d, 128.25d, 128.78d, 129.23d, 132.45s,
136.27s, 138.21s, 139.96s, 143.97s. Anal. Calcd for
C26H31NO: C, 83.60; H, 8.37; N, 3.75. Found: C, 83.79;
H, 8.51; N, 3.72.
Acknowledgements
This work was supported in part by a grant from the
Ministry of Education, Culture, Sports, Science and
Technology of Japan and the promotion and Mutual Aid
Corporation for Private Schools of Japan.
References and notes
1. For reviews on organometallic additions to imine and their
derivatives, see: (a) Enders, D.; Reinhold, U. Tetrahedron:
Asymmetry 1997, 8, 1895–1946. (b) Bloch, R. Chem. Rev. 1998,
98, 1407–1438. (c) Alvaro, G.; Savoia, D. Synlett 2002,
651–673. (d) Martelli, G.; Savoia, D. Curr. Org. Chem. 2003,
7, 1049–1070.
2. (a) Takahashi, H.; Hsieh, B. C.; Higashiyama, K. Chem. Pharm.
Bull. 1990, 38, 2429–2434. (b) Higashiyama, K.; Inoue, H.;
Yamauchi, T.; Takahashi, H. J. Chem. Soc., Perkin Trans. 1
1995, 111–115. (c) Yamauchi, T.; Takahashi, H.; Higashiyama,
K. Chem. Pharm. Bull. 1998, 46, 384–389. (d) Yamauchi, T.;
Takahashi, H.; Higashiyama, K. Heterocycles 1998, 48,
1813–1823. (e) Ishii, A.; Miyamoto, F.; Higashiyama, K.;
Mikami, K. Tetrahedron Lett. 1998, 39, 1199–1202. (f)
Poerwono, H.; Higashiyama, K.; Takahashi, H. J. Org. Chem.
1998, 63, 2711–2714.
3. (a) Maruoka, K.; Imoto, H.; Saito, S.; Yamamoto, H. J. Am.
Chem. Soc. 1994, 116, 4131–4132. (b) Maruoka, K.; Ito, M.;
Yamamoto, H. J. Am. Chem. Soc. 1995, 117, 9091–9092. (c)
Ooi, T.; Miura, T.; Kondo, Y.; Maruoka, K. Tetrahedron Lett.
1997, 38, 3947–3950. (d) Ooi, T.; Kondo, Y.; Kon-I, K.;
Maruoka, K. Chem. Lett. 1998, 403–404. (e) Saito, S.; Sone, T.;
Shimada, K.; Yamamoto, H. Synlett 1999, 81–83. (f) Marx, A.;
Yamamoto, H. Synlett 1999, 584–586. (g) Saito, S.; Shiozawa,
M.; Yamamoto, H. Angew. Chem., Int. Ed. 1999, 38,
1769–1771. (h) Saito, S.; Yamazaki, S.; Yamamoto, H.
Angew. Chem., Int. Ed. 2001, 40, 3613–3617. (i) Saito, S.;
Nagahara, M.; Shiozawa, M.; Nakadai, H.; Yamamoto, H.
J. Am. Chem. Soc. 2003, 125, 6200–6210. (j) Ito, H.; Nagahara,
T.; Ishihara, K.; Saito, S.; Yamamoto, H. Angew. Chem., Int.
Ed. 2004, 43, 994–997.
4.3. General procedure for removal of the N-2,4,6-
trimethylbenzyl group from (R,R)- and (S,R)-2e
A single diastereomer of (R,R)- and (S,R)-2e (0.134 mmol)
and trifluoroacetic acid (5 mL) was stirred at 50 8C for 3
days, and then diluted with water (20 mL). The resulting
aqueous phase was basified with 10% NaOH solution and
extracted with CH2Cl2 (3!20 mL). The combined extracts
were washed with brine, dried over anhydrous Na2SO4, and
concentrated under reduced pressure to leave a oily residue,
which was subjected to column chromatography on silica
gel with ethyl acetate–hexane (1:2) to give (R,R)- and (S,R)-
3, respectively.
4. (a) Denmark, S.; Seierstad, M. J. Org. Chem. 1999, 64,
1610–1619. (b) Kanemasa, S.; Ueno, N.; Shirahase, M.
Tetrahedron Lett. 2002, 43, 657–660.
5. (a) Takahashi, H.; Suzuki, Y.; Kametani, T. Heterocycles 1983,
20, 607–610. (b) Takahashi, H.; Niwa, H.; Higashiyama, K.
Heterocycles 1988, 27, 2099–2102. (c) Pridgen, L. N. J. Org.
Chem. 1991, 56, 1340–1344. (d) Higashiyama, K.; Inoue, H.;
Takahashi, H. Tetrahedron 1994, 50, 1083–1092. (e) Poerwono,
H.; Higashiyama, K.; Yamauchi, T.; Takahashi, H. Hetero-
cycles 1997, 46, 385–400. (f) Poerwono, H.; Higashiyama, K.;
Yamauchi, T.; Kubo, H.; Ohmiya, S.; Takahashi, H.
Tetrahedron 1998, 54, 13955–13970.
4.3.1. Synthesis of (Z,R)-N-isopropyliminium–aluminum
complex (4h) for NOE experiment. A mixture of 2,6-
diphenylphenol (0.148 g, 0.6 mmol) and Me3Al (0.1 mL,
0.2 mmol; 2 M in toluene) in dry CH2Cl2 was stirred at rt
under nitrogen for 30 min to afford the solution of ATPH.6
Then the solution was concentrated under reduced pressure
and residue was solved in CDCl3 (0.5 mL). To this solution
was added the solution of oxazolidine (1c) (0.1 mmol) in
CDCl3 (0.5 mL) and stirred at rt for 2 h to obtain the CDCl3
solution of iminium–aluminum complex (4h). The solution
was used for NOE experiment without purification and the
data was showed some peak for complicated chart: 1H NMR
(CDCl3) d: 0.55 (3H, brd), 0.75 (3H, brd), 1.41 (3H, br),
1.54 (1H, br), 2.67 (1H, br), 2.99 (1H, br), 3.50 (1H, brdd),
6.52 (2H, brd).
6. The assignment of 1H NMR spectrum on 10 and 20 position of 4h
determined as follows. 20-Dideuteriumed iminium–aluminum
complex (4h0) was obtained from (R)-phenylglycine with
LiAlD4 for 5 steps by the similar procedure.4 In 4h0, the signals
1
of H NMR spectrum on 20 position (d: 1.54, 2.67 ppm) were
not observed, and the signal on 10 position (d: 3.50 ppm) was
observed. Therefore, the assignment of 1H NMR spectrum on 10
and 20 position of 4h was determined.