Chiral o-methoxyaryldiazaphosphonamides - A new class of efficient Lewis bases in the catalytic asymmetric ring opening of cyclooctene oxide with silicon tetrachloride

Article abstract of DOI:10.1002/1099-0690(200108)2001:15<2819::AID-EJOC2819>3.0.CO;2-R

The synthesis of a new class of chiral o-methoxydiazaphosphonamides (as Lewis bases) has been investigated, together with their use as catalysts in the asymmetric ring opening (ARO) of cyclooctene oxide with silicon tetrachloride. Enantiomeric excesses varying from 6 to 99% ee were observed, depending on the nature of the catalyst examined. On the basis of experimental considerations, a mechanistic rationale involving hexacoordinate silicon species has been proposed.

Full text of DOI:10.1002/1099-0690(200108)2001:15<2819::AID-EJOC2819>3.0.CO;2-R

FULL PAPER
Chiral o-Methoxyaryldiazaphosphonamides ؊ A New Class of Efficient Lewis
Bases in the Catalytic Asymmetric Ring Opening of Cyclooctene Oxide with
Silicon Tetrachloride
[a]
[a]
[a]
[a]
´
´
Sebastien Reymond, Olivier Legrand, Jean Michel Brunel, and Gerard Buono*
Keywords: Asymmetrization / o-Methoxyaryldiazaphosphonamides / Lewis bases / Ring opening / meso-Epoxides
The synthesis of a new class of chiral o-methoxydiazaphos- omeric excesses varying from 6 to 99% ee were observed,
phonamides (as Lewis bases) has been investigated, together
depending on the nature of the catalyst examined. On the
with their use as catalysts in the asymmetric ring opening basis of experimental considerations, a mechanistic rationale
(ARO) of cyclooctene oxide with silicon tetrachloride. Enanti-
involving hexacoordinate silicon species has been proposed.
Introduction
In this paper we describe the synthesis of various new
organophosphorus Lewis bases and their successful use in
the asymmetrization of cyclooctene oxide. We also report
evidence in support of a mechanistic hypothesis implying
hexacoordinate cationic silicon intermediates.
In the last few years, the asymmetric ring opening of ep-
oxides (ARO) has appeared to constitute one of the most
efficient methods for the preparation of various complex
molecules.[1] Numerous procedures for asymmetrization of
meso compounds have been developed; these include depro-
tonation,[2] protonation,[3] esterification,[4] and hydrolysis.[5]
Although, in asymmetric synthesis, the use of nucleophiles
such as halide ion (Cl, Br, F) is one method of choice for
the preparation of valuable intermediates bearing two con-
tiguous stereogenic centers, no efficient, catalytic, enantio-
selective method has yet been reported. Thus, although
the classical addition of hydrogen chloride or hydrogen
bromide remains as the simplest, most atom-efficient
method, the need to manipulate sensitive polyfunctional in-
termediates has resulted in the introduction of a variety of
new and milder alternatives for this transformation.[6] Only
Denmark et al., in 1998, have described the formation of
optically active chlorohydrins in moderate enantiomeric ex-
cesses (ee values of up to 87%), through the use of chiral
phosphoramide Lewis bases as catalysts, activating various
silylated nucleophiles.[7]
Results and Discussion
The synthesis of various diastereomerically pure o-me-
thoxydiazaphosphonamides Lewis bases has been achieved
by methylation of the corresponding o-hydroxyarylphos-
phonamides, easily prepared by a stereospecific and regiose-
lective 1,3 migration rearrangement reaction as outlined in
Scheme 2.[9,10] Methylation of these compounds produced
the expected pure o-methoxydiazaphosphonamides Lewis
bases, obtained in diastereomerically pure form and in high
chemical yields, varying from 84 to 91% (Scheme 3). These
Lewis bases have been successfully employed in the catalytic
asymmetric ring opening of cyclooctene oxide as test sub-
strate, with freshly distilled SiCl4 in CH2Cl2 at Ϫ78 °C
(Table 1).[11]
Recently we reported a beneficial secondary ligand inter-
action effect in the asymmetric ring opening of cyclooctene
oxide with SiCl4, catalyzed by a chiral o-methoxyphenyldi-
azaphosphonamide Lewis base 9, and producing ee values of
up to 99% (Scheme 1).[8]
In all cases, the chlorohydrin was obtained in good yields,
varying from 60 to 83%, but the enantioselectivity of the
reaction was highly dependent on the catalyst structure.
Thus, depending on the nature of the substituent on the
aryl group, a significant decrease in the ee was sometimes
encountered, values varying from 99 to 6%. Moreover, the
lower ee values were observed when using catalyst 13 and
15, affording the expected chlorohydrin with only 20 and
6% ee, respectively. In the case of catalyst 13, this important
lack of enantioselectivity can be explained by a decrease in
the donor character of the methoxy group, and con-
sequently by a weak coordination of this latter with the
silicon atom. An interesting point to mention is that the use
of diastereomers anti-9 and syn-10 produced the same re-
sults in terms of enantioselectivities and that the chlorohyd-
rins in both cases possess the same absolute configuration
(1R,2R).[12]
Scheme 1. Catalytic asymmetric ring opening of meso-epoxides
[a]
´
`
´ ´
Ecole Nationale Superieure de Syntheses, de Procedes et
´
d’Ingenierie Chimiques d’Aix Marseille, UMR CNRS 6516,
´
´ ˆ
Faculte de St Jerome,
Av. Escadrille Normandie Niemen, 13397 Marseille, Cedex 20,
France
Fax: (internat.) ϩ 33-4/91027776
E-mail: buono@spi-chim.u-3mrs.fr
Eur. J. Org. Chem. 2001, 2819Ϫ2823
WILEY-VCH Verlag GmbH, 69451 Weinheim, 2001
1434Ϫ193X/01/0815Ϫ2819 $ 17.50ϩ.50/0
2819

S. Reymond, O. Legrand, J. M. Brunel, G. Buono
FULL PAPER
Scheme 2. Synthesis of chiral o-hydroxyaryldiazaphosphonamides 1؊8
Scheme 3. Synthesis of chiral o-methoxyaryldiazaphosphonamides 9؊16
In order to explain all the experimental results, a mechan- now well known that NLE represents an important tool for
istic rationale involving hexacoordinate silicon species has the explanation of mechanisms, providing useful insights
been proposed.[13] Thus, as a test of this assumption, non- both into the behavior of enantioselective catalytic systems
linear effect (NLE) studies were performed. Indeed, it is and into the mechanisms of the processes they mediate.[14]
2820
Eur. J. Org. Chem. 2001, 2819Ϫ2823

Chiral o-Methoxyaryldiazaphosphonamides Ϫ A New Class of Efficient Lewis Bases
FULL PAPER
Table 1. Catalytic asymmetric ring opening of cyclooctene oxide
with SiCl4
Entry[a]
Lewis base
Solvent
Yield (%)[c]
ee (%)[d]
Scheme 4. Catalyst disproportionation
1
2[b]
3
4
5
6
7
8
9
9
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
THF
77
60
75
70
83
72
77
65
69
Ͼ 99
34
Ͼ 99
88
Ͼ 99
20
95
9
10
11
12
13
14
15
16
Conclusion
In conclusion, we have developed the synthesis of a new
class of diastereomerically pure o-methoxydiazaphosphona-
mides Lewis bases and their successful use in the enantiose-
lective, catalytic ring opening of cyclooctene oxide with
SiCl4, with ee values of up to 99% in numerous cases. Fur-
ther studies dealing with application of these catalysts in
other asymmetric reactions are currently under investi-
gation.
6
Ͼ 99
[a]
Reactions performed on 1.2 mmol scale at Ϫ 78 °C over 4 h,
using freshly distilled SiCl4. Ϫ
mercial SiCl4 without any purification. Ϫ
flash chromatography. Ϫ
[b]
Reaction performed using com-
[c]
Isolated yield after
ee values determined by GC analysis,
[d]
using a LIPODEX E column, the major enantiomer possessing
(1R,2R) absolute configuration.
Experimental Section
Consequently, experiments were performed to determine
whether NLEs were operative in the enantioselective cata-
lyzed ring opening of meso-epoxides with SiCl4. A strong
positive NLE was encountered, suggesting the existence of
various catalytic species (Figure 1).
1
Materials and Methods: H, 13C, 19F, and 31P NMR spectra were
recorded with Bruker AC 100 and AC 200 spectrometers, in CDCl3
as solvent. The chemical shifts (ppm) were determined relative to
Me4Si (1H and 13C) and 85% H3PO4 (31P). Ϫ Optical rotations
were measured with a PerkinϪElmer 241 MC polarimeter. Ϫ Tetra-
hydrofuran (THF) was distilled from sodium/benzophenone ketyl
immediately prior to use. Dichloromethane was distilled from P2O5.
Silicon tetrachloride was distilled under argon immediately prior
to use. Ethyl acetate and petroleum ether (35Ϫ60 °C) were pur-
chased from SDS and used without any previous purification. Ϫ
Column chromatography was performed on SDS silica gel (70Ϫ230
mesh). Ϫ Enantiomeric excesses of the chlorohydrin were deter-
mined by GC analysis after derivatization of the chlorohydrin using
trifluoroacetic anhydride (Lipodex E, 30 KPa, isothermal). Ϫ Pre-
cursors 1؊8 were prepared as previously described.[9]
General Procedure for the Preparation of Catalysts 9؊16: K2CO3
(400 mg, 2.9 mmol) was added at 25 °C under argon to a solution
of precursor 1؊8 (0.67 mmol) in CH2Cl2 (30 mL) and MeOH
(10 mL), followed by dropwise addition of MeI (1 mL, 16 mmol).
The mixture was stirred under reflux for 12 h. After removal of the
solvent in vacuo, CH2Cl2 (20 mL) was added to the residue, and
the mixture was filtered through a small pad of silica gel and con-
centrated in vacuo. Purification by silica gel chromatography
yielded the desired compounds 9؊16.
Figure 1. NLEs observed in the asymmetric ring opening of cyclo-
(2S,5S)-2-(2-Anisyl)-3-phenyl-1,3-diaza-2-phosphabicyclo-
[3.3.0]octane 2-Oxide (9): This compound was obtained by follow-
ing the general procedure with precursor 1 {(2S,5S)-2-(2-hydroxy-
phenyl)-3-phenyl-1,3-diaza-2-phosphabicyclo[3.3.0]octane 2-oxide}
(210 mg, 0.67 mmol). Purification by silica gel chromatography
(ethyl acetate) afforded 185 mg (84%) of compound 9 as a white
solid, m.p. 144 °C. Ϫ [α]2D0 ϭ Ϫ28.9 (c ϭ 1.0, CH2Cl2). Ϫ 31P NMR
(40.5 MHz, CDCl3): δ ϭ 24.4. Ϫ 1H NMR (200 MHz, CDCl3):
δ ϭ 8.06 (ddd, J ϭ 14.9, 7.4, 2.2 Hz, 1 H), 7.38Ϫ7.32 (m, 1 H),
octene oxide with SiCl4 using catalyst 9
Moreover, on the basis of these results we may postulate
that catalyst disproportionation had occurred and that the
formation of a stable [Cl3Si(RL*)(SL*)] cationic complex
18 results in a catalytically inactive reservoir for the minor
(RL*) ligand, in equilibrium with the active species
[Cl3Si(SL*)] (Scheme 4).[13a,15]
Eur. J. Org. Chem. 2001, 2819Ϫ2823
2821

S. Reymond, O. Legrand, J. M. Brunel, G. Buono
FULL PAPER
7.16Ϫ6.93 (m, 5 H), 6.84Ϫ6.71 (m, 2 H), 4.14Ϫ3.72 (m, 6 H), 3.47 8.7 Hz), 55.5, 29.6, 28.9 (d, J ϭ 11.0 Hz), 28.5 (d, J ϭ 8.3), 28.3
(td, J ϭ 8.2, 3.3 Hz, 1 H), 3.04Ϫ2.87 (m, 1 H), 2.17Ϫ1.73 (m, 4 (d, J ϭ 5.8), 24.8, 24.7. Ϫ C15H23N2O2P (294.33): calcd. C 61.2, H
H). Ϫ 13C NMR (50 MHz, CDCl3): δ ϭ 160.5, 142.1 (d, J ϭ 7.9, N 9.5, P 10.5; found C 61.5, H 7.7, N 9.4, P 10.6.
6.6 Hz), 136.6 (d, J ϭ 7.3 Hz), 133.5, 128.8 (2 C), 120.6, 120.3,
(2S,5S)-2-(5-Chloro-2-methoxyphenyl)-3-phenyl-1,3-diaza-2-phos-
120.0 (d, J ϭ 163.0 Hz), 115.8 (d, J ϭ 4.4 Hz), 111.0, 110.9, 59.3
phabicyclo[3.3.0]octane 2-Oxide (13): This compound was obtained
(d, J ϭ 6.2 Hz), 55.8, 48.8 (d, J ϭ 15.9 Hz), 45.1, 33.4, 26.5 (d, J ϭ
by following the general procedure from precursor 5 {(2S,5S)-2-
2.5 Hz). Ϫ C18H21N2O2P (328.35): calcd. C 65.8, H 6.5, N 8.5, P
(5-chloro-2-hydroxyphenyl)-3-phenyl-1,3-diaza-2-phosphabicyclo-
9.4; found C 65.5, H 6.6, N 8.4, P 9.5.
[3.3.0]-octane 2-oxide} (348.5 mg, 0.67 mmol). Purification by silica
(2R,5S)-2-(2-Anisyl)-3-phenyl-1,3-diaza-2-phosphabicyclo-
[3.3.0]octane 2-Oxide (10): This compound was obtained by follow-
ing the general procedure with precursor 2 {(2R,5S)-2-(2-hydroxy-
phenyl)-3-phenyl-1,3-diaza-2-phosphabicyclo[3.3.0]octane 2-oxide}
(210 mg, 0.67 mmol). Purification by silica gel chromatography
(ethyl acetate) afforded 189 mg (86%) of compound 10 as a white
solid, m.p. 198 °C. Ϫ [α]2D0 ϭ ϩ108.7 (c ϭ 1.15, CH2Cl2). Ϫ 31P
NMR (40.5 MHz, CDCl3): δ ϭ 18.9. Ϫ 1H NMR (200 MHz,
CDCl3): δ ϭ 8.13 (ddd, J ϭ 14.9, 7.4, 2.2 Hz, 1 H), 7.40Ϫ7.32 (m,
1 H), 7.09Ϫ6.95 (m, 5 H), 6.75Ϫ6.66 (m, 2 H), 4.28Ϫ4.18 (m, 1
H), 3.85Ϫ3.65 (m, 1 H), 3.62 (s, 3 H), 3.55Ϫ3.47 (m, 1 H),
3.06Ϫ2.93 (m, 1 H), 2.79Ϫ2.64 (m, 1 H), 2.16Ϫ2.05 (m, 1 H),
2.01Ϫ1.81 (m, 2 H), 1.67Ϫ1.54 (m, 1 H). Ϫ 13C NMR (50 MHz,
CDCl3): δ ϭ 159.9 (d, J ϭ 2.4 Hz), 142.6 (d, J ϭ 5.9 Hz), 138.9
(d, J ϭ 4.8 Hz), 134.2, 128.6 (2 C), 120.9, 120.6, 120.0, 115.1 (d,
J ϭ 118 Hz), 110.3 (d, J ϭ 6.2 Hz), 58.4 (d, J ϭ 9.1 Hz), 53.5 (d,
J ϭ 9.4 Hz), 50.0, 44.7 (d, J ϭ 6.4 Hz), 31.5 (d, J ϭ 3.7 Hz), 27.7
(d, J ϭ 5.4 Hz). Ϫ C18H21N2O2P (328.35): calcd. C 65.8, H 6.5, N
8.5, P 9.4; found C 65.6, H 6.8, N 8.6, P 9.4.
gel chromatography (ethyl acetate/petroleum ether, 80:20) afforded
219 mg (90%) of compound 13 as a white solid, m.p. 171 °C. Ϫ
[α]2D0 ϭ Ϫ9.5 (c ϭ 0.105, CH2Cl2). Ϫ 31P NMR (40.5 MHz,
1
CDCl3): δ ϭ 22.4. Ϫ H NMR (200 MHz, CDCl3): δ ϭ 8.07 (dd,
J ϭ 15.1, 2.8 Hz, 1 H), 7.29 (dd, J ϭ 15.1, 2.8 Hz, 1 H), 7.14 (dd,
J ϭ 7.1, 7.6 Hz, 2 H), 7.02 (d, J ϭ 7.6 Hz, 2 H), 6.83 (t, J ϭ 7.1 Hz,
1 H), 6.65 (dd, J ϭ 8.8, 6.9 Hz, 1 H), 3.96Ϫ3.75 (m, 3 H), 3.69 (s,
3 H), 3.52Ϫ3.42 (m, 1 H), 3.00Ϫ2.87 (m, 1 H), 2.15Ϫ1.67 (m, 4
H). Ϫ 13C NMR (50 MHz, CDCl3): δ ϭ 158.9, 141.6 (d, J ϭ
6.7 Hz), 136.1 (d, J ϭ 8.7 Hz), 133.1, 129.0, 125.6 (d, J ϭ 17.9 Hz),
122.3 (d, J ϭ 162.6 Hz), 121.0, 115.8 (d, J ϭ 4.4 Hz), 112.3 (d, J ϭ
9.8 Hz), 59.3 (d, J ϭ 6.2 Hz), 56.2, 48.7 (d, J ϭ 15.9 Hz), 44.9,
33.3, 26.4. Ϫ C18H20ClN2O2P (362.79): calcd. C 59.6, H 5.6, N 7.7,
P 8.5; found C 59.7, H 5.6, N 7.4, P 8.9.
(2S,5S)-2-(2,5-Dimethoxyphenyl)-3-phenyl-1,3-diaza-2-phosphabi-
cyclo[3.3.0]octane 2-Oxide (14): This compound was obtained by
following the general procedure from precursor 6 {(2S,5S)-2-(2-hy-
droxy-5-methoxyphenyl)-3-phenyl-1,3-diaza-2-phosphabicyclo-
[3.3.0]octane 2-oxide} (230.5 mg, 0.67 mmol). Purification by silica
gel chromatography (ethyl acetate/petroleum ether, 80:20) afforded
214 mg (89%) of compound 14 as a white solid, m.p. 164 °C. Ϫ
[α]2D0 ϭ Ϫ28.7 (c ϭ 0.115, CH2Cl2). Ϫ 31P NMR (40.5 MHz,
(2S,5S)-2-(5-Fluoro-2-methoxyphenyl)-3-phenyl-1,3-diaza-2-phos-
phabicyclo[3.3.0]octane 2-Oxide (11): This compound was obtained
by following the general procedure from precursor 3 {(2S,5S)-2-(5-
fluoro-2-hydroxyphenyl)-3-phenyl-1,3-diaza-2-
1
CDCl3): δ ϭ 24.0. Ϫ H NMR (200 MHz, CDCl3): δ ϭ 7.62 (dd,
phosphabicyclo[3.3.0]octane 2-oxide} (222.4 mg, 0.67 mmol). Puri-
fication by silica gel chromatography (ethyl acetate/petroleum ether,
80:20) afforded 211 mg (91%) of compound 11 as a white solid,
m.p. 160 °C. Ϫ [α]2D0 ϭ Ϫ64.3 (c ϭ 0.375, CH2Cl2). Ϫ 19F NMR
(94.2 MHz, CDCl3): δ ϭ Ϫ68.5. Ϫ 31P NMR (40.5 MHz, CDCl3):
δ ϭ 24.0. Ϫ 1H NMR (200 MHz, CDCl3): δ ϭ 7.91 (ddd, J ϭ 15.6,
8.4, 2.3 Hz, 1 H), 7.28Ϫ7.09 (m, 5 H), 6.96Ϫ6.89 (m, 1 H),
6.80Ϫ6.72 (m, 1 H), 4.20Ϫ3.87 (m, 3 H), 3.79 (s, 3 H), 3.64Ϫ3.53
(m, 1 H), 3.00Ϫ3.16 (m, 1 H), 2.26Ϫ2.13 (m, 2 H), 2.10Ϫ1.82 (m,
2 H). Ϫ 13C NMR (50 MHz, CDCl3): δ ϭ 156.7 (dd, J ϭ 240.6,
19.8 Hz), 156.5, 141.7 (d, J ϭ 7.2 Hz), 128.9 (2 C), 122.9 (dd, J ϭ
24.5, 8.6 Hz), 120.9, 119.6 (d, J ϭ 23.3 Hz), 115.8, 115.9, 112.1 (dd,
J ϭ 10.0, 7.3 Hz), 122.1 (dd, J ϭ 163, 5.7 Hz), 59.3 (d, J ϭ 5.9 Hz),
56.4, 48.8 (d, J ϭ 15.9 Hz), 45.0, 33.3, 26.4. Ϫ C18H20FN2O2P
(346.34): calcd. C 62.4, H 5.8, N 8.1, P 8.9; found C 62.0, H 5.7,
N 8.1, P 9.0.
J ϭ 16.0, 3.2 Hz, 1 H), 7.01 (d, J ϭ 8.0 Hz, 1 H), 7.09 (dd, J ϭ
8.8, 8.0 Hz, 1 H), 6.89 (dd, J ϭ 8.9, 3.2 Hz, 1 H), 6.79 (t, J ϭ
7.0 Hz, 1 H), 6.66 (t, J ϭ 8.8 Hz, 1 H), 4.0Ϫ3.69 (m, 6 H), 3.64 (s,
3 H), 3.45 (td, J ϭ 8.3, 3.2 Hz, 1 H), 3.01Ϫ2.90 (m, 1 H), 2.15Ϫ1.67
(m, 4 H). Ϫ 13C NMR (50 MHz, CDCl3): δ ϭ 154.7, 153.2 (d, J ϭ
17.2 Hz), 141.9 (d, J ϭ 7.2 Hz), 128.8 (2 C), 121.1 (d, J ϭ
152.9 Hz), 120.6, 120.5, 119.4, 115.8, 115.7, 112.4 (d, J ϭ 10.2 Hz),
59.2 (d, J ϭ 5.8 Hz), 56.4, 55.9, 48.9, 48.6, 45.0, 33.3, 26.4 (d, J ϭ
3.2 Hz). Ϫ C19H23N2O3P (358.37): calcd. C 63.7, H 6.5, N 7.8, P
8.6; found C 64.0, H 6.6, N 7.6, P 8.6.
(2S,5S)-2-(2-Methoxybiphenyl-3-yl)-3-phenyl-1,3-diaza-2-phos-
phabicyclo[3.3.0]octane 2-Oxide (15): This compound was obtained
by following the general procedure from precursor 7 {(2S,5S)-2-(2-
hydroxybiphenyl-3-yl)-3-phenyl-1,3-diaza-2-phosphabicyclo[3.3.0]-
octane 2-oxide} (261.3 mg, 0.67 mmol). Purification by silica gel
chromatography (ethyl acetate) afforded 238 mg (88%) of com-
pound 15 as a white solid, m.p. 112 °C. Ϫ [α]2D0 ϭ Ϫ33.0 (c ϭ 0.225,
(1R,6R)-8-(2-Anisyl)-7,9-dimethyl-7,9-diaza-8-phosphabi-
cyclo[4.3.0]nonane 8-Oxide (12): This compound was obtained by
following the general procedure from precursor 4 {(1R,6R)-8-(2-
hydroxyphenyl)-7,9-dimethyl-7,9-diaza-8-phosphabicyclo-[4.3.0]-
nonane 8-oxide} (188 mg, 0.67 mmol). Purification by silica gel
chromatography (ethyl acetate) afforded 179 mg (91%) of com-
pound 12 as a white solid, m.p. 115 °C. Ϫ [α]2D0 ϭ Ϫ28.3 (c ϭ 1.0,
1
CH2Cl2). Ϫ 31P NMR (40.5 MHz, CDCl3): δ ϭ 24.4. Ϫ H NMR
(200 MHz, CDCl3): δ ϭ 8.01 (ddd, J ϭ 14.7, 7.2, 2.4 Hz, 1 H),
7.48Ϫ7.09 (m, 11 H), 6.87Ϫ6.80 (m, 1 H), 4.04Ϫ3.86 (m, 3 H),
3.52Ϫ3.44 (m, 1 H), 3.24 (s, 3 H), 3.15Ϫ2.85 (m, 1 H), 2.13Ϫ1.75
(m, 4 H). Ϫ 13C NMR (50 MHz, CDCl3): δ ϭ 159.3, 142.1 (d, J ϭ
5.7 Hz), 138.0, 133.5, 135.1, 135.0, 134.7 (d,
J ϭ 7.1 Hz),
1
CH2Cl2). Ϫ 31P NMR (40.5 MHz, CDCl3): δ ϭ 32.2. Ϫ H NMR
129.0Ϫ128.4 (m, 5 C), 127.4, 126.4 (d, J ϭ 132.4 Hz), 123.9, 123.6,
120.7, 115.9 (d, J ϭ 4.1 Hz), 60.3, 58.9 (d, J ϭ 6.2 Hz), 48.8 (d,
J ϭ 15.8 Hz), 45.0, 32.9, 26.4. Ϫ C24H25N2O2P (404.44): calcd. C
71.3, H 6.2, N 6.9, P 7.7; found C 71.4, H 6.5, N 6.8, P 7.9.
(200 MHz, CDCl3): δ ϭ 8.00 (ddd, J ϭ 13.9, 7.6, 1.8 Hz, 1 H),
7.49Ϫ7.37 (m, 1 H), 6.98 (td, J ϭ 7.4, 2.6 Hz, 1 H), 6.87 (dd, J ϭ
8.2, 6.0 Hz, 1 H), 3.89 (s, 3 H), 2.80Ϫ3.00 (m, 2 H), 2.45 (d, J ϭ
12 Hz, 2 H), 2.35 (d, J ϭ 11.6 Hz, 2 H), 2.10Ϫ1.85 (m, 4 H),
1.5Ϫ1.2 (m, 4 H). Ϫ 13C NMR (50 MHz, CDCl3): δ ϭ 160.9, 138.0 (2S,5S)-2-(1-Methoxynaphth-2-yl)-3-phenyl-1,3-diaza-2-phosphabi-
(d, J ϭ 5.9 Hz), 133.7, 120.5 (d, J ϭ 12.9 Hz), 118.0 (d, J ϭ
152.0 Hz), 111.0 (d, J ϭ 7.7 Hz), 64.8 (d, J ϭ 5.8 Hz), 64.0 (d, J ϭ
cyclo[3.3.0]octane 2-Oxide (16): This compound was obtained by
following the general procedure from precursor 8 {(2S,5S)-2-(1-hy-
2822
Eur. J. Org. Chem. 2001, 2819Ϫ2823

Chiral o-Methoxyaryldiazaphosphonamides Ϫ A New Class of Efficient Lewis Bases
FULL PAPER
[5]
M. Kurihara, J. Kamiyama, S. Kobayashi, M. Ohno, Tetrahed-
ron Lett. 1985, 26, 5831.
H. Masumada, K. Takase, M. Nishio, A. Hasegawa, Y. Nishi-
yama, Y. Ishii, J. Org. Chem. 1994, 59, 5550 and references
therein.
S. E. Denmark, P. A. Barsanti, K. T. Wong, R. A. Stavenger,
J. Org. Chem. 1998, 63, 2428.
droxynaphth-2-yl)-3-phenyl-1,3-diaza-2-phosphabicyclo[3.3.0]-
octane 2-oxide} (245.2 mg, 0.67 mmol). Purification by silica gel
chromatography (ethyl acetate) afforded 213 mg (84%) of com-
pound 16 as a white solid, m.p. 92 °C. Ϫ [α]2D0 ϭ Ϫ165 (c ϭ 0.3,
[6]
1
CH2Cl2). Ϫ 31P NMR (40.5 MHz, CDCl3): δ ϭ 23.9. Ϫ H NMR
[7]
[8]
[9]
(200 MHz, CDCl3): δ ϭ 7.99Ϫ7.93 (m, 2 H), 7.77Ϫ7.72 (m, 1 H),
7.59Ϫ7.53 (m, 1 H), 7.46Ϫ7.35 (m, 2 H), 7.09Ϫ6.97 (m, 4 H),
6.76Ϫ6.68 (m, 1 H); 4.09Ϫ3.74 (m, 6 H), 3.53Ϫ3.43 (m, 1 H),
2.98Ϫ2.82 (m, 1 H), 2.11Ϫ1.69 (m, 4 H). Ϫ 13C NMR (50 MHz,
CDCl3): δ ϭ 159.5, 142.0, 141.8, 137.1, 130.2 (d, J ϭ 8.4 Hz), 128.2
(2 C), 127.9 (d, J ϭ 12.2 Hz), 127.3 (d, J ϭ 3.8 Hz), 126.0, 123.8,
123.6, 122.9, 121.6 (d, J ϭ 164.9 Hz), 120.6, 115.7 (d, J ϭ 4.6 Hz),
62.7, 59.1 (d, J ϭ 6.2 Hz), 48.8 (d, J ϭ 15.9 Hz), 44.82, 32.7, 26.4.
Ϫ C22H23N2O2P (378.40): calcd. C 69.8, H 6.1, N 7.4, P 8.2; found
C 69.7, H 6.4, N 7.3, P 9.0.
J. M. Brunel, O. Legrand, S. Reymond, G. Buono, Angew.
Chem. Int. Ed. 2000, 39, 2554Ϫ2557.
For the synthesis of chiral o-hydroxyarylphosphane oxides, see:
[9a]
O. Legrand, J. M. Brunel, T. Constantieux, G. Buono,
[9b]
Chem. Eur. J. 1998, 4, 1061Ϫ1067. Ϫ
O. Legrand, J. M.
[9c]
Brunel, G. Buono, Eur. J. Org. Chem. 1999, 1099Ϫ1105. Ϫ
O. Legrand, J. M. Brunel, G. Buono, Tetrahedron 2000, 56,
595Ϫ603.
[10]
For applications of o-hydroxyarylphosphane oxides in asym-
metric catalysis, see: [10a] J. M. Brunel, T. Constantieux, O. Leg-
rand, G. Buono, Tetrahedron Lett. 1998, 39, 2961Ϫ2964. Ϫ [10b]
O. Legrand, J. M. Brunel, G. Buono, Tetrahedron Lett. 1998,
39, 9419Ϫ9422. Ϫ
General Procedure for the Catalytic Asymmetric Ring Opening of
Cyclooctene Oxide (Table 1, Entry 1): SiCl4 (140 µL, 1.22 mmol)
was added at Ϫ78 °C under argon to a stirred solution of catalyst
9 (0.122 mmol) in CH2Cl2 or THF (6 mL). After 5 min, cyclooc-
tene oxide (154 mg, 1.22 mmol) was added. After the addition was
complete, the mixture was stirred at Ϫ78 °C for 4 h and then
quenched by pouring into cold (Ϫ78 °C), rapidly stirred saturated
KF/KH2PO4 (1:1) (15 mL). It was then allowed to warm to room
temperature. The resulting mixture was extracted with CH2Cl2 (2
ϫ 15 mL) and the combined organic fractions were washed with
water (10 mL), dried with MgSO4, filtered, and concentrated under
reduced pressure. The residue was purified by silica gel chromato-
[10c]
J. M. Brunel, O. Legrand, G. Buono,
[10d]
Tetrahedron: Asymmetry 1999, 10, 1979Ϫ1984. Ϫ
J. M.
Brunel, O. Legrand, G. Buono, Eur. J. Org. Chem. 2000,
3313Ϫ3321.
[11]
HCl is well known to open epoxides easily. Because of this fact,
it was essential to establish that the SiCl4 used was free of HCl
in order to obtain reproducible results: A. D. Cross, Quart. Rev.
Chem. Soc. 1960, 14, 317.
At this stage of our knowledge about this reaction, we are not
able to explain this fact clearly and experiments are in due
course in our laboratory.
[12]
[13]
Cationic silicon species have been already postulated as inter-
mediates in other Lewis base promoted reactions, see: [13a]S. E.
graphy (petroleum ether/diethyl ether, 90:10), affording (1R,2R)-2-
20
chlorocyclooctan-1-ol in 77% yield. Ϫ [α]
ϭ Ϫ1.4 (c ϭ 1.0,
Denmark, X. Su, Y. Nishigaichi, J. Am. Chem. Soc. 1998, 120,
365
[13b]
CH2Cl2); ee Ͼ 99% (determined by GC analysis after derivatization
of the chlorohydrin using trifluoroacetic anhydride), tR(1R,2R) ϭ
31.4 min (Ͼ 99%), tR(1S,2S) ϭ 51.2 min (Ͻ 1%) (Lipodex E, 30
12990. Ϫ
A. R. Bassindale, J. C. Y. Lau, P. G. Taylor, J.
[13c]
Organomet. Chem. 1995, 499, 137. Ϫ
A. R. Bassindale, J.
C. Y. Lau, P. G. Taylor, J. Organomet. Chem. 1995, 499, 75.
For a recent review on nonlinear effects in asymmetric syn-
thesis, see: C. Girard, H. B. Kagan, Angew. Chem. Int. Ed.
1998, 37, 2922Ϫ2959.
[14]
[15]
1
KPa, isothermal 120 °C, 0.1 µL). Ϫ H NMR (200 MHz, CDCl3):
δ ϭ 4.11 (ddd, J ϭ 9.2, 7.3, 2.8 Hz, 1 H), 3.86 (ddd, J ϭ 9.2, 6.9,
2.1 Hz, 1 H), 2.63 (br s, 1 H), 2.29Ϫ2.13 (m, 1 H), 2.07Ϫ1.30 (m,
11 H). Ϫ 13C NMR (50 MHz, CDCl3): δ ϭ 76.2, 71.2, 32.3, 32.0,
25.7, 25.7, 24.9, 24.1. Ϫ C8H15ClO (162.66): calcd. C 59.1, H 9.3;
found C 58.9, H 9.2.
All these hypothesis support the existence of the already pro-
posed catalytic cycle involving pentacoordinate and hexacoord-
inate cationic silicon species (Scheme 5).[8]
Acknowledgments
We thank the CNRS for its financial support. O. L. acknowledges
the CNRS and Region PACA for a doctoral fellowship. S. R.
acknowledges MENRT for a doctoral fellowship.
[1] [1a]
R. A. Johnson, K. B. Sharpless, in: Catalytic Asymmetric
Synthesis (Ed.: I. Ojima), VCH, New York, 1993, chapter 4.1.
[1b]
Ϫ
E. N. Jacobsen, in: Catalytic Asymmetric Synthesis (Ed.:
I. Ojima), VCH, New York, 1993, chapter 4.2. Ϫ [1c] A. S. Rao,
S. K. Paknikar, J. G. Kirtane, Tetrahedron 1983, 39, 2323. Ϫ
[1d]
E. N. Jacobsen, in: Comprehensive Asymmetric Catalysis
(Eds.: E. N. Jacobsen, A. Pfaltz, H. Yamamoto), Springer-Ver-
[1e]
lag, Berlin, 1999, vol. III, p. 1309. Ϫ
D. M. Hodgson, A.
R. Gibbs, G. P. Lee, Tetrahedron 1996, 52, 14361Ϫ14384. Ϫ [1f]
I. Sakurada, S. Yamasaki, R. Gottlich, T. Iida, M. Kanai, M.
Shibasaki, J. Am. Chem. Soc. 2000, 122, 1245Ϫ1246. Ϫ
D. Nugent, J. Am. Chem. Soc. 1998, 120, 7139Ϫ7140.
[1g]
W.
Scheme 5. Proposed mechanistic rationale for the enantioselective
ring opening of cyclooctene oxide
[2]
[3]
N. S. Simpkins, J. Chem. Soc., Chem. Commun. 1986, 88Ϫ89.
K. Fuji, M. Node, S. Terada, M. Murata, H. Nagasawa, J. Am.
Chem. Soc. 1985, 107, 6404Ϫ6405.
[4]
[4a] T. Mukaiyama, I. Tomioka, M. Shimizu, Chem. Lett. 1984,
Received August 17, 2000
(publication delayed at the author’s request)
[O00436]
[4b]
49Ϫ50. Ϫ
J. Hiratake, K. Shibata, N. Baba, J. Oda, Syn-
thesis 1988, 278.
Eur. J. Org. Chem. 2001, 2819Ϫ2823
2823