Stereoselective synthesis of syn- and anti-1,2-aminoalcohols using iridium-catalyzed allylic amination reactions

Article abstract of DOI:10.1055/s-0029-1217814

A methodology for the stereoselective synthesis of syn- and anti-1,2-aminoalcohols, employing iridium-catalyzed amination reactions of allylic carbonates, was developed. Georg Thieme Verlag Stuttgart.

Full text of DOI:10.1055/s-0029-1217814

LETTER
2281
Stereoselective Synthesis of syn- and anti-1,2-Aminoalcohols Using Iridium-
Catalyzed Allylic Amination Reactions
Stereoselective
Sy
o
nthesis of syn
s
-
and anti
h
-1,2-Aminoa
i
lcohol
y
s
asu Ichikawa,* Shun-Ichi Yamamoto, Hiyoshizo Kotsuki, Keiji Nakano
Faculty of Science, Kochi University, Akebono-cho, Kochi 780-8520, Japan
Fax +81(88)8448359; E-mail: ichikawa@kochi-u.ac.jp
Received 8 April 2009
branched allylamines.5 The regioselectivity of this pro-
Abstract: A methodology for the stereoselective synthesis of syn-
and anti-1,2-aminoalcohols, employing iridium-catalyzed amina-
tion reactions of allylic carbonates, was developed.
cess complements those typically observed in palladium-
catalyzed allylic aminations. Hartwig developed an ele-
gant variant of this reaction, which involves iridium-cata-
lyzed asymmetric allylic amination6 employing Feringa’s
phosphoramidite ligands.7 Although extensive studies of
this amination reaction have been conducted,8 reactions of
chiral allylic carbonates with branching alkoxy groups a
to the allyl unit, leading to the formation of diastereomer-
ically pure 1,2-amino alcohols, have not been explored.
Key words: allylic amination, amino alcohol, catalysis, iridium,
transition metals
Over the past several decades, chiral 1,2-aminoalcohols
have become an increasingly important group of com-
pounds owing to their high synthetic versatility. These
substances serve as useful intermediates in the synthesis
of several members of natural product families, as well as
chiral auxiliaries and ligands in asymmetric synthesis.1
For example, the 1,2-aminoalcohol moiety is found in
polyoxamic acid (1, Figure 1), which is a structural con-
stituent of the nucleoside antibiotic polyoxin,2 as well as
in indolizine alkaloids such as 1,2-dihydroxyindolizine
(2).3 Although several useful methods exist to prepare op-
tically active 1,2-aminoalcohol building blocks, more ef-
ficient and practical routes are still needed.
At the outset of this investigation, we prepared the simple
model allylic carbonate 6, bearing silyloxy-branching a to
the allylic carbonate unit, starting from methyl (S)-(–)-lac-
tate (3) (Scheme 1).9 We chose the tert-butyldimethylsilyl
(TBS) oxygen protecting group owing to its wide popular-
ity and usefulness in natural product synthesis.
O
O
TBSCl
imidazole
1) DIBAL-H, CH2Cl2
OMe
O
OMe
DMF, r.t.
(quant)
2) (EtO)2P(O)CH2CO2Et
NaH, THF
OH
OTBS
3
(74%)
4
OH
OH NH2
HO
O
HO
OH
1) DIBAL-H, CH2Cl2
OEt
O
OMe
2) ClCO2Me, pyridine
CH2Cl2
N
OH
O
OTBS
OTBS
1,2-dihydroxyindolizine (2)
5
6
polyoxamic acid (1)
(93%)
Scheme 1 Synthesis of model substrate 6
Figure 1 Representative natural products possessing chiral 1,2-
aminoalcohol units
With this test substrate in hand, the iridium-catalyzed al-
lylic amination reaction using benzylamine in THF was
explored (Scheme 2).10 The iridium catalyst was activated
by heating [Ir(cod)Cl]2, ligand L1 (Figure 2) and propyl-
amine in THF at 50 °C to induce cyclometalation.11 Initial
experiments under several conditions varying amounts of
benzylamine and catalyst and a range of temperatures
demonstrated that the reaction was sluggish with most of
the starting material being recovered. Continual efforts re-
vealed that the use of DMF as solvent led to a dramatic ac-
celeration of the reaction rate without sacrificing the
regio- and stereoselectivity. The reaction using 4 mol% of
[Ir(cod)Cl]2, 8 mol% of L1 and 1.5 equiv of benzylamine
in DMF at room temperature was completed in 1.5 days
(Scheme 2). In this process, the branched product 7 was
formed predominantly (7/9 = 90:10) and with a high dia-
stereoselectivity (8 was not observed by 1H NMR analy-
sis). However, the observation that yields were
Recent investigations in our group have revealed a proce-
dure for the stereoselective synthesis of allylic 1,2-ami-
noalcohols that is based upon [3,3]-sigmatropic
rearrangement of allyl cyanates derived from chiral a-
alkoxy aldehydes.4 While this methodology has unques-
tionable synthetic utility, it requires several steps to con-
struct the 1,2-aminoalcohol motif. In this letter, we
describe a more concise and efficient procedure to rapidly
access both syn- and anti-1,2-aminoalcohols that is based
upon iridium-catalyzed allylic amination reactions.
In 2001, Takeuchi reported the amination reactions of ter-
minal allylic carbonates using iridium complexes to form
SYNLETT 2009, No. 14, pp 2281–2286
x
x
.x
x
.2
0
0
9
Advanced online publication: 12.08.2009
DOI: 10.1055/s-0029-1217814; Art ID: U03609ST
© Georg Thieme Verlag Stuttgart · New York

2282
Y. Ichikawa et al.
LETTER
consistently hovering around 50% indicated that a more dimethoxyaniline a potentially more useful nucleophile
reactive nitrogen nucleophile was needed.
than p-anisidine in the iridium-catalyzed amination reac-
tion.
O
1) [Ir(cod)Cl]2 (4 mol%), L1 (8 mol%)
Table 1 Iridium-Catalyzed Allylic Amination Using p-Anisidine
and 2,4-Dimethoxyaniline
PrNH2, THF, activation
O
OMe
2) PhCH2NH2, DMF
r.t.
O
6
O
1) [Ir(cod)Cl]2 (4 mol%)
TBS
N
L1 or L2 (8 mol%), activation
O
OMe
OTBS
2)
R
OMe
PhH2C
H
H
CH2Ph
H
O
CH2Ph
DMF, r.t.
N
N
6
H2N
(1.5 equiv)
TBS
O
TBS
O
TBS
8
9
7
R
OMe
R
OMe MeO
R
Scheme 2 Iridium-catalyzed allylic amination using benzylamine
HN
HN
NH
Ph
Ph
OTBS
OTBS
TBSO
10a: R = H
10b: R = OMe
11a: R = H
11b: R = OMe
12a: R = H
12b: R = OMe
O
O
P
N
P
N
O
O
Entrya
Ligand
L1
R
H
Yield (%)b
anti/sync
93:7
b/ld
Ph
Ph
A
B
C
D
89
93
95
94
94:6
98:2
96:4
97:3
(aR,R,R)-L1
(aS,S,S)-L2
L2
H
8:92
92:8
2:98
Figure 2 Phosphoramidite ligands L1 and L2 used to generate
active catalysts
L1
OMe
OMe
L2
Iridium-promoted amination reaction of 6 using p-anisi-
dine as the nucleophile was then explored.12 It was expect-
ed that the products of this process could be readily
transformed into the corresponding free amines through
oxidative removal of the N-aryl group.13 We did not ex-
pect an improvement of this reaction owing to the de-
creased basicity of the aryl amine. Surprisingly, the
reaction of allylic carbonate 6 with 1.5 equiv of p-anisi-
dine and the iridium complex prepared from 4 mol% of
[Ir(cod)Cl]2 and 8 mol% of L1 at room temperature in
DMF proceeded smoothly and was complete in one day
(Table 1, entry A).14 A high branched-to-linear product
ratio was observed (b/l = 94:6) and the branched products,
comprised of the anti- and syn-1,2-aminoalcohols 10a and
11a, respectively, were isolated in 89% yield with high
anti-selectivity (10a/11a = 93:7). When the ligand L2
was employed in this process under identical conditions
(entry B), the branched products were produced in 93%
yield with high syn-selectivity (11a/10a = 92:8). This ob-
servation indicates that catalyst-control of stereochemis-
try is dominant in this reaction.15
a Catalyst {L1, L2 and [Ir(cod)Cl]2} was pre-activated with n-propyl-
amine.
b Yield of the isolated branched allylamines.
c The ratio was determined by HPLC analysis of isolated branched
allylamines.
d Ratio of branched/linear product.
The stereochemistry of the 1,2-aminoalcohols generated
in these reactions were initially assigned by the empirical
rule reported by Hartwig.6 To confirm the assignments of
stereochemistry, the aminoalcohols were converted into
the corresponding oxazolidinones (Scheme 3).18 Treat-
ment of syn-aminoalcohol 11b with CAN in a mixture of
acetonitrile and pH 7 phosphate buffer, led to formation of
the corresponding primary amine, which was subsequent-
ly reacted with methyl chloroformate and aqueous sodium
bicarbonate to afford methyl carbamate 13 in 75–82%
yields.19 Silyl deprotection with tetrabutylammonium fluo-
ride produced a mixture of 14 and 15, which was then
treated with sodium hydride in THF to form the trans-
oxazolidinone 15. Similar transformation of anti-1,2-
aminoalcohol 10b afford the cis-oxazolidinone 16. Both
H-1 and H-2 protons of the trans-oxazolidinone 15 reso-
nate at higher fields than those of the cis-stereoisomer 16
(Figure 3). In addition, steric interaction between the me-
thyl and vinyl groups in the cis-oxazolidinone 16 results
in a g-effect on 13C NMR spectra, which produces an up-
field shift of the methyl carbon resonance. These observa-
tions show that the relative stereochemistry of 15 and 16
are trans and cis, respectively.
Reactions employing 2,4-dimethoxyaniline were ex-
plored because the products were expected to undergo N-
dearylation more efficiently than their counterparts pre-
pared from p-anisidine.16 Although steric hindrance asso-
ciated with the o-methoxy group in 2,4-dimethoxyaniline
was a concern, the reaction was found to take place with
similar yields and selectivities to those observed with p-
anisidine (entries C and D).17 These results, coupled with
an expected facile N-deprotection protocol, made 2,4-
Synlett 2009, No. 14, 2281–2286 © Thieme Stuttgart · New York

LETTER
Stereoselective Synthesis of syn- and anti-1,2-Aminoalcohols
2283
MeO
HN
OMe
1) CAN
MeCN, phosphate buffer
possible intermediates for the preparation of polyoxamic
acid (1) and 1,2-dihydroxyindolizine (2). In the synthesis
of the polyoxamic acid precursor A,20 a-hydroxy ester 18
was prepared from L-ascorbic acid (17)21 and then protect-
ed as its TBS ether 19a (Scheme 5). A four-step sequence,
involving reduction with DIBAL-H, Wittig olefination, a
second DIBAL-H reduction, and treatment with methyl
chloroformate, then furnished the allylic carbonate 20a.
O
HN
OMe
O
2)
aq. NaHCO3
Cl
OMe
TBSO
OTBS
(75–82%)
13
11b
O
HN
OMe
NaH, THF
(42%)
OH
n-Bu4NF, MeCN
O
NH
O
O
O
HO
O
1) ref. 20
O
HO
O
15
OMe
14
2) TBSCl or TESCl
imidazole
OR
18: R = H
19a: R = TBS (93%)
19b: R = TES (82%)
HO
OH
steps
L-ascorbic acid (17)
10b
O
NH
1) DIBAL-H, CH2Cl2
2) Ph3PCHCO2Et
O
16
O
O
O
Scheme 3 Transformation of 1,2-aminoalcohols into the correspon-
ding oxazolidinones
3) DIBAL-H, CH2Cl2
4) ClCO2Me, pyridine, CH2Cl2
O
OMe
RO
20a: R = TBS (77%)
20b: R = TES (60%)
δ
C 15.9
Me
H1
δ
C 18.9
Me
H1
Scheme 5 Preparation of allylic carbonate for the synthesis of poly-
oxamic acid
H2
H2
δH 3.90
δH 4.30
δ
H 4.33
δ
H 4.84
O
O
NH
J1,2 = 7.7 Hz
NH
J1,2 = 7.2 Hz
Interestingly, iridium-catalyzed amination of allylic car-
bonate 20a was quite sluggish even under forcing condi-
tions (Table 2). For example, reaction of 20a, using 12.5
mol% iridium catalyst and four equiv of 2,4-dimethoxy-
aniline in DMF at room temperature for one week, led to
only 30% conversion (determined by 1H NMR analysis).
Experimentation led to the finding that changing the pro-
tecting group from TBS to triethylsilyl (TES) led to an in-
crease in the rate of this reaction (Table 2, entry A). TES-
protected substrate 20b was prepared (Scheme 5,
18→19b→20b) and then subjected to iridium-catalyzed
amination. The amination reaction of 20b using L1 was
complete in one week and the branched products were ef-
ficiently generated (70% yield and b/l = 87:13) favoring
the syn-product 21 (dr 94:6). To check the catalyst-con-
trolled stereochemistry, allylic amination using L2 was
carried out (entry B). Although a considerable amount of
linear product 23 was produced (b/l = 42:58), the anti-1,2-
aminoalcohol 22 was formed as the major product (dr
83:17).
O
O
15
16
Figure 3 Comparison of chemical shifts of trans- and cis-oxazoli-
dinones
MeO
HN
OMe
OH NH2
O
HO
OH
O
OH
O
OP
1
A
OP
OH
O
HO
HN
N
O
2
B
MeO
OMe
Scheme 4 Intermediates for the synthesis of polyoxamic acid and
1,2-dihydroxyindolizine
The synthesis of the anti-1,2-aminoalcohol B required for
synthesis of 1,2-dihydroindolizine began with the forma-
tion of 25 from D-isoascorbic acid (24) using a known
procedure (Scheme 6).22 Implementation of the transfor-
mations shown in Scheme 5, converted 25 into the sub-
strate 26, which was suitable for iridium-catalyzed
amination reaction.
As demonstrated by the results presented in Table 1, the
choice of ligand L1 versus L2 in the iridium-catalyzed al-
lylic amination reactions enables selective synthesis of ei-
ther syn- or anti-1,2-aminoalcohols with high degrees of
stereoselectivity. Armed with these results, we next fo-
cused our attention on the synthesis of syn- and anti-1,2-
aminoalcohols A and B (Scheme 4), which could serve as
Synlett 2009, No. 14, 2281–2286 © Thieme Stuttgart · New York

2284
Y. Ichikawa et al.
LETTER
Table 2 Synthesis of Polyoxamic Acid Intermediate
Table 3 Synthesis of 1,2-Dihydroxyindolizine Intermediate
1) [Ir(cod)Cl]2 (12.5 mol%)
L1 or L2 (25 mol%), activation
1) [Ir(cod)Cl]2 (12.5 mol%)
L2 or L1 (25 mol%), activation
O
O
O
O
O
O
MeO
OMe
2) MeO
OMe
DMF, r.t.
2)
O
OMe
O
OMe
DMF, r.t.
OTES
OR
H2N
H2N
26
20a: R = TBS
20b: R = TES
(4.0 equiv)
(4.0 equiv)
R
R
R
R
O
HN
O
HN
O
HN
O
HN
O
O
O
O
O
OTES
OTES
OTES
OTES
22
28
21
27
O
O
MeO
OMe
MeO
R =
OMe
O
R
R
R =
N
N
H
H
OTES
OTES
29
23
Entrya
Ligand
L2
Yield (%)b
anti/sync
90:10
b/ld
Entrya
Ligand
L1
Yield (%)b
syn/antic
94:6
b/ld
A
B
58
70
73:27
A
B
70
36
87:13
42:58
L1
16:84
>99:<1
L2
17:83
a Catalyst {L1, L2 and [Ir(cod)Cl]2} was pre-activated with n-propyl-
amine.
a Catalyst {L1, L2 and [Ir(cod)Cl]2} was pre-activated with n-propyl-
amine.
b Yield of the isolated branched allylamines 27 and 28.
c The ratio was determined by HPLC analysis of isolated branched
allylamines.
b Yield of the isolated branched allylamines 21 and 22.
c The ratio was determined by HPLC analysis of isolated branched
allylamines.
d Ratio of branched/linear product.
d Ratio of branched/linear product.
In summary, we have investigated iridium-catalyzed ami-
nation reactions of a-silyloxy-containing allylic carbon-
ates in the context of stereoselective methods for the
synthesis of anti- and syn-1,2-aminoalcohols. The reac-
tions were found to proceed smoothly in DMF, and 2,4-
dimethoxyaniline proved to be a competent nucleophile.
As a result, the resulting products can be readily trans-
formed into free amines by oxidative removal of the N-
aryl group. Although somewhat lower yields and regio-
selectivities were observed in the reactions of some com-
plex substrates (e.g. 20b and 26), catalyst-controlled
stereoselection was also attained in these processes. This
is the first example of catalyst-control of diastereoselec-
tivity in the iridium-catalyzed allylic amination of chiral
allylic carbonates. Further efforts are underway to im-
prove this process.
OH
O
O
HO
O
ref. 21
O
CO2Et
HO
OH
OTES
25
D-isoascorbic acid (24)
1) DIBAL-H, CH2Cl2
O
O
O
O
O
OMe
2)
Cl
OMe
OTES
pyridine, CH2Cl2
(86%)
26
Scheme 6 Preparation of allylic carbonate for the synthesis of
1,2-dihydroxyindolizine
Construction of the anti-1,2-aminoalcohol moiety in B
was carried out by employing the reaction of allylic car-
bonate 26 and 2,4-dimethoxyaniline in the presence of the
iridium catalyst prepared from ligand L2 (Table 3). The
reaction, which was complete in one week, afforded the
branched products in a 58% yield (b/l = 73:27). HPLC
analysis of the branched products showed that the desired
anti-product 27 was formed with 90:10 stereoselectivity.
In contrast, reaction of 26 using ligand L1 led to the gen-
eration of the branched products exclusively in a 70%
yield with a dr of 84:16 favoring the syn-isomer 28.
Acknowledgment
We are grateful for financial support provided by Scientific Re-
search on Priority Areas (18032055 and 18037053) and a Grant-in-
Aid for Scientific Research (17310129) from MEXT. This work
was supported in part by a Special Research Grant for Green Sci-
ence from Kochi University.
References and Notes
(1) (a) Ager, D. J.; Prakash, I.; Scheed, D. R. Chem. Rev. 1996,
96, 835. (b) Bergmeier, S. C. Tetrahedron 2000, 56, 2561.
Synlett 2009, No. 14, 2281–2286 © Thieme Stuttgart · New York

LETTER
Stereoselective Synthesis of syn- and anti-1,2-Aminoalcohols
2285
(2) For isolation, see: (a) Isono, K.; Asahi, K.; Suzuki, S. J. Am.
Chem. Soc. 1969, 91, 7499. For synthetic works, see:
(b) Saksena, A. K.; Lovey, R. G.; Girijavallabhan, V. M.;
Ganguly, A. K. J. Org. Chem. 1986, 51, 5024. (c) Hirama,
M.; Hioki, H.; Ito, S. Tetrahedron Lett. 1988, 29, 3125.
(d) Garner, P.; Park, J. M. J. Org. Chem. 1988, 53, 2979.
(e) Dureult, A.; Carreaux, F.; Depezay, J.-C. Tetrahedron
Lett. 1989, 30, 4527. (f) Paz, M.; Sardina, F. J. Org. Chem.
1993, 58, 6990. (g) Dondoni, A.; Franco, S.; Merchan, F. L.;
Merino, P.; Tejero, T. Tetrahedron Lett. 1993, 34, 5479.
(h) Marshall, J. A.; Seletsky, B. M.; Coan, P. S. J. Org.
Chem. 1994, 59, 5139. (i) Matsura, F.; Hamada, Y.; Shioiri,
T. Tetrahedron Lett. 1994, 35, 733. (j) Jackson, R. F.;
Palmer, N. J.; Wyther, M. J.; Clegg, W.; Elsegood, M.
J. Org. Chem. 1995, 60, 6431. (k) Kang, S. H.; Choi, H.
Chem. Commun. 1996, 1521. (l) Veeresa, G.; Datta, A.
Tetrahedron Lett. 1998, 39, 119. (m) Savage, I.; Thomas,
E. J.; Wilson, P. D. J. Chem. Soc., Perkin Trans. 1 1999,
3291. (n) Davis, F. A.; Prasad, K. R.; Carroll, P. J. J. Org.
Chem. 2002, 67, 7802. (o) Tarrade, A.; Dauban, P.; Dodd,
R. H. J. Org. Chem. 2002, 68, 9521.
(3) (a) Harris, T. M.; Harris, C. M. H.; Hill, J. E.; Ungemach,
F. S.; Broquist, H. P.; Wickwire, B. M. J. Org. Chem. 1987,
52, 3094. (b) Heitz, M.-P.; Overman, L. E. J. Org. Chem.
1989, 54, 2591.
(4) (a) Ichikawa, Y.; Ito, T.; Isobe, M. Chem. Eur. J. 2005, 11,
1949. (b) Ichikawa, Y.; Matsunaga, K.; Masuda, T.;
Kotsuki, H.; Nakano, K. Tetrahedron 2008, 64, 11313.
(5) (a) Takeuchi, R.; Ue, N.; Tanabe, K.; Yamashita, K.; Shiga,
N. J. Am. Chem. Soc. 2001, 123, 9525. For the allylic
alkylation, see: (b) Takeuchi, R.; Kashio, M. Angew. Chem.
Int. Ed. Engl. 1997, 36, 263. (c) Takeuchi, R.; Kashio, M.
J. Am. Chem. Soc. 1998, 120, 8647. For reviews, see:
(d) Takeuchi, R. Synlett 2002, 1954. (e) Takeuchi, R.;
Kezuka, S. Synthesis 2006, 3349.
(6) Ohmura, T.; Hartwig, J. F. J. Am. Chem. Soc. 2002, 15164.
(7) (a) Feringa, B. L. Acc. Chem. Res. 2000, 346. For a
convenient synthetic method for L1 and L2 ligands, see:
(b) Alexakis, A.; Gille, S.; Prian, F.; Rosset, S.; Ditrich, K.
Tetrahedron Lett. 2004, 45, 1449. (c) Polet, D.; Alexakis,
A. Org. Lett. 2005, 7, 1621.
(8) (a) Lipowsky, G.; Helmchen, G. Chem. Commun. 2004,
116. (b) Welter, C.; Koch, O.; Lipowsky, G.; Helmchen, G.
Chem. Commun. 2004, 896. (c) Helmchen, G. Chem.
Commun. 2004, 116. (d) Weihofen, R.; Dahnz, A.;
Tverskoy, O.; Helmchen, G. Chem. Commun. 2005, 3541.
(e) Weihofen, R.; Tverskoy, O.; Helmchen, G. Angew.
Chem. Int. Ed. 2006, 45, 5546. (f) Lee, J. H.; Shin, S.; Kang,
J.; Lee, S. J. Org. Chem. 2007, 72, 7443.
(14) Representative Experimental Procedure: A Schlenk flask
under argon was charged with [Ir(cod)Cl]2 (21 mg, 0.032
mmol) and L1 (32 mg, 0.060 mmol). THF (3.0 mL) and n-
propylamine (3.0 mL) were added, and the reaction mixture
was stirred at 50 °C for 30 min. Evaporation of the volatile
materials gave the activated catalyst as a crude yellow solid,
which was dissolved in DMF (2.0 mL) and used as catalyst
for the next reaction. Under an argon atmosphere, allylic
carbonate 6 (205 mg, 0.75 mmol) and 2,4-dimethoxyaniline
(175 mg, 1.10 mmol) were added to the solution quickly, and
the flask was sealed under argon. After standing at room
temperature for 1 d, the reaction mixture was diluted with
H2O and Et2O. The organic layer was separated and the
aqueous layer was extracted with Et2O. The combined
organic extracts were washed with H2O and brine, dried over
Na2SO4, filtered and concentrated under reduced pressure.
Purification by silica gel chromatography (Et2O–hexane,
1:10) gave an inseparable mixture of branched allylamines
10b and 11b (250 mg, 95%) and linear allylamine 12b (10
mg, 4%). The resulting branched allylamines were analyzed
by HPLC to determine the ratio to be 92:8. Branched
allylamine 10b: 1H NMR (C6D6, 400 MHz): d = 0.06 (s, 3
H), 0.13 (s, 3 H), 1.01 (s, 9 H), 1.12 (d, J = 6.5 Hz, 3 H), 3.31
(s, 3 H), 3.46 (s, 3 H), 3.63 (m, 1 H), 4.00 (dq, J = 6.5, 3.5
Hz, 1 H), 4.73 (d, J = 8.0 Hz, 1H, NH), 5.13 (ddd, J = 10.5,
2.0, 1.0 Hz, 1 H), 5.19 (ddd, J = 17.5, 2.0, 1.0 Hz, 1 H), 5.85
(ddd, J = 17.5, 10.5, 7.0 Hz, 1 H), 6.47 (dd, J = 8.5, 2.5 Hz,
1 H), 6.51 (d, J = 2.5 Hz, 1 H), 6.64 (d, J = 8.5 Hz, 1 H); 13
C
NMR (CDCl3, 100 MHz): d = –5.08, –4.16, 18.0, 20.6, 25.7,
55.3, 55.7, 62.5, 70.6, 99.1, 103.6, 111.8, 117.4, 131.5,
136.3, 148.3, 151.7.
(15) The allylic amination of 6 with an achiral iridium complex
decorated with triphenylphosphite was briefly investigated.
In this case, a 53:47 mixture of products 10b and 11b,
together with recovered starting material 6 (17%) was
isolated (77% yield based on the consumed starting material;
Scheme 7).
[Ir(cod)Cl]2 (10 mol%)
P(OPh)3 (40 mol%)
O
10b and 11b
(53:47)
MeO
OMe
OTBS
O
OMe
(1.5 equiv)
6
H2N
EtOH, 50 °C, 20 h
Scheme 7
(16) Our initial attempts to promote N-dearylation of 11a with
CAN were complicated by formation of varying amounts of
p-quinone ii (Scheme 8). See also the reference 18b.
(9) Ichikawa, Y.; Tsuboi, K.; Isobe, M. J. Chem. Soc., Perkin
Trans. 1 1994, 2791.
O
OMe
(10) Hartwig reported that solvent influences the reactivity and
enantioselectivity of enantioselective allylic amination.
Although reactions in the polar solvents, such as DMF and
EtOH were fast (100% conversion after 2 h), low ee’s were
observed (80–77% ee). As a result, THF was recommended
as the most suitable balance of rate (100% conversion after
8–10 h) and enantioselectivity (95% ee). See reference 6.
(11) (a) Kiener, C. A.; Shu, C.; Incarvito, C.; Hartwig, J. F. J. Am.
Chem. Soc. 2003, 125, 14272. (b) Leitner, A.; Shu, C.;
Hartwig, J. F. Org. Lett. 2005, 7, 1093.
(12) Shu, C.; Leitner, A.; Hartwig, J. F. Angew. Chem. Int. Ed.
2004, 43, 4797.
(13) (a) Fukuyama, T.; Frank, R. K.; Jewell, C. F. J. Am. Chem.
Soc. 1980, 102, 2122. (b) Kronenthal, D. R.; Han, C. Y.;
Taylor, M. K. J. Org. Chem. 1982, 47, 2765.
N
NH2
HN
CAN
MeCN–H2O
+
OTBS
OTBS
OTBS
i
ii
11a
Scheme 8
(17) A competitive experiment using a 1:1 mixture of p-anisidine
and 2,4-dimethoxyaniline was carried out in order to
compare reactivity. In this reaction, a 7:3 mixture of
products 10b and 10a was obtained in 86% combined yield
(Scheme 9). This indicates that 2,4-dimethoxyaniline is
approximately two times more nucleophilic than p-anisi-
dine.
Synlett 2009, No. 14, 2281–2286 © Thieme Stuttgart · New York

2286
Y. Ichikawa et al.
LETTER
r.t. for 50 min, the reaction mixture was diluted with H2O
and Et2O. The separated aqueous layer was extracted with
Et2O, and the combined organic extracts were washed with
aqueous 1M KHSO4, H2O, saturated aqueous NaHCO3 and
brine, dried over Na2SO4, filtered and concentrated under
reduced pressure. Purification by silica gel chromatography
(EtOAc–hexane, 1:5) gave methyl carbamate 13 (138 mg,
82% from 11b). 1H NMR (CDCl3, 400 MHz): d = 0.03 (s, 3
H), 0.05 (s, 3 H), 0.87 (s, 9 H), 1.16 (d, J = 6.0 Hz, 3 H), 3.69
(s, 3 H), 3.92 (m, 1 H), 4.05 (m, 1H), 5.02 (br, 1 H, NH), 5.14
(dt, J = 10.5, 1.5 Hz, 1 H), 5.19 (dt, J = 17.0, 1.5 Hz, 1 H),
5.81 (ddd, J = 17.0, 10.5, 5.5 Hz, 1 H); 13C NMR (CDCl3,
100 MHz): d = –4.90, –4.48, 17.9, 20.7, 25.7, 52.0, 58.6,
70.0, 115.2, 137.2, 156.8.
O
1) [Ir(cod)Cl]2 (4 mol%)
L1 (8 mol%), activation
O
OMe
OTBS
2)
MeO
OMe
OMe
6
H2N
H2N
DMF, r.t.
(1:1)
MeO
OMe
HN
OMe
HN
OTBS
OTBS
(20) In initial attempts to synthesize the intermediate in the
polyoxamic acid synthesis, allylic carbonate i was
10b
10a
Scheme 9
synthesized from an intermediate prepared in our previous
synthesis of polyoxamic acid (ref. 4a). Iridium-catalyzed
allylic amination of i resulted in predominant formation of
the linear product ii, and none of desired branched products
was recognized (Scheme 10). It appears that the acetonide
group is responsible for the erosion in regioselectivity
through steric interactions and/or chelation effects. Further
studies to reveal the effects of protecting groups on the
iridium-catalyzed allylic amination reaction are underway.
(18) For the empirical rule for assignment of relative confi-
gurations of trans- and cis-oxazolidinones, see:
(a) Futagawa, S.; Inui, T.; Shiba, T. Bull. Chem. Soc. Jpn.
1973, 46, 3308. (b) Murakami, M.; Ito, H.; Ito, Y. J. Org.
Chem. 1993, 58, 6766. Small difference in J1,2 values
between 15 (7.2 Hz) and 16 (7.7 Hz) may not allow the use
of empirical rule reported by Futagawa. We have noticed a
similar example to determine the relative configuration of
oxazolidinones i and ii reported by Kim: (c) Kim, J. D.;
Zee, O. P.; Jung, Y. H. J. Org. Chem. 2003, 68, 3721; In this
case, the g-effect of the 13C NMR spectra (dC = 19.6 ppm for
i and 16.8 ppm for ii) may support the assignment of relative
stereochemistry (Figure 4).
1) [Ir(cod)Cl]2 (4 mol%)
TBSO
L1 (8 mol%), activation
O
O
OMe
O
2)
H2N
OMe
O
O
DMF, 60 °C, 1 d
δ
H 4.08
δ
H 4.43
δ
H 4.88
δ
H 4.47
H1
O
H2
i
H1
O
H2
Me
Ph
TBSO
OMe
Me
Ph
NH
NH
N
H
O
J1,2 = 7.4 Hz
J1,2 = 7.7 Hz
(45%)
O
O
ii
i
ii
Figure 4
Scheme 10
(19) Oxidative N-Dearylation: To a solution of branched
allylamine 11b (220 mg, 0.61 mmol) in a mixture of MeCN
(10 mL) and pH 7 phosphate buffer (10 mL) at 0 °C, was
added CAN (1.10 g, 2.46 mmol). After stirring at 0 °C for 15
min, the reaction mixture was diluted with saturated aqueous
NaHCO3 (10 mL) and then treated with methyl
(21) Wei, C. C.; Bernardo, S. D.; Tengi, J. P.; Borgese, J.;
Weigele, M. J. Org. Chem. 1985, 50, 3462.
(22) (a) Cohen, N.; Banner, B. L.; Laurenzano, A. J.; Carozza, L.
Org. Synth. Coll. Vol. 7; John Wiley & Sons: London, 1990,
297. (b) Hinman, A.; Du Bois, J. J. Am. Chem. Soc. 2003,
125, 11510.
chloroformate (0.10 mL, 1.3 mmol) at 0 °C. After stirring at
Synlett 2009, No. 14, 2281–2286 © Thieme Stuttgart · New York