Lewis acid promoted rearrangements of 1,3-dioxolanyl-substituted 1,2-oxazines into novel products with 1,3,6-trioxa-7-azacyclopenta[cd]indene skeletons

Article abstract of DOI:10.1055/s-2006-956463

Lewis acid promoted rearrangements of 4-methoxy- and 4-benzyloxy- substituted 1,2-oxazines syn-1b and syn-1c furnished novel tricyclic products 5 and 6. A mechanistic rationale is suggested for the different rearrangement pathways depending on the configu

Full text of DOI:10.1055/s-2006-956463

3498
LETTER
Lewis Acid Promoted Rearrangements of 1,3-Dioxolanyl-Substituted 1,2-
Oxazines into Novel Products with 1,3,6-Trioxa-7-azacyclopenta[cd]indene
Skeletons
R
earrangemen
a
ts of 1,3-D
b
i
bstitute
a
d 1,2-Oxaz
n
ines Pfrengle, Ahmed Al-Harrasi, Hans-Ulrich Reissig*
Institut für Chemie und Biochemie, Freie Universität Berlin, Takustraße 3, 14195 Berlin, Germany
Fax +49(30)83855367; E-mail: hans.reissig@chemie.fu-berlin.de
Received 29 September 2006
Dedicated to Professor Manfred Christl on the occasion of his 65th birthday
1,2-Oxazines anti-1 and syn-1 are easily available in a ste-
Abstract: Lewis acid promoted rearrangements of 4-methoxy- and
4-benzyloxy-substituted 1,2-oxazines syn-1b and syn-1c furnished
novel tricyclic products 5 and 6. A mechanistic rationale is suggest-
ed for the different rearrangement pathways depending on the con-
reodivergent manner by reaction of lithiated alkoxyal-
lenes with a D-glyceraldehyde-derived nitrone.2 When 4-
methoxy-substituted 1,2-oxazine anti-1b was exposed to
figuration and the nature of the 4-alkoxy groups of the precursor tin tetrachloride the tricyclic acetal 3 was obtained in 50%
1,2-oxazines. Short period hydrogenolyses of these rearrangement
products afforded tetrahydrofuranyl-annulated 5,6-dihydro-4H-1,2-
oxazines 10 and 11, whereas longer reduction times led to formation
derivative of bicyclic 1,2-oxazin-4-one 2. Constitution
of tetrahydrofuran derivatives 14, 15 and 16, 17 in good yields.
yield rather than the expected product 2 (Scheme 1).
Compound 3 can be regarded as an internally protected
and configuration of 3 were proven by X-ray crystallo-
graphic analysis.3
Key words: heterocycles, 1,2-oxazines, furans, Lewis acids, rear-
rangements, 1,2-alkyl shifts, hydrogenations
The Lewis acid induced reaction of 4-methoxy-substitut-
ed 1,2-oxazine syn-1b followed a different, even more
surprising pathway giving the novel product 5 with three
annulated heterocycles (hexahydro-2H,4H-1,3,6-trioxa-
We recently reported that enantiopure 4-(2-trimethylsi-
lyl)ethoxy-substituted 1,2-oxazines such as anti-1a or
7-azacyclopenta[cd]indene derivative) in good yield
syn-1a smoothly undergo Lewis acid promoted rearrange-
(Scheme 2).4 This transformation was also performed
ments providing bicyclic ketones 2 or 4 (Scheme 1 and
with 4-benzyloxy-substituted precursor syn-1c which fur-
nished the analogous tricyclic product 6 in 55% yield.
Scheme 2).1 These bicyclic 1,2-oxazine derivatives are
versatile precursors for the synthesis of highly substituted
tetrahydropyrans which can be regarded as carbohydrate
mimetics. We now present the results of experiments in-
volving related 1,2-oxazines with other 4-alkoxy substitu-
ents, which led to different types of rearrangement
products.
O
O
OTBS
H
H
R = TMSE
N
OR
O
O
Lewis acid
ref. 1
O
Bn
O
4
SnCl4, MeCN,
–30 °C to r.t., 8–9 h,
O
O
H
N
OTBS
Bn
O
H
H
R = TMSE syn-1a
O
R = TMSE
Lewis acid
R = Me
R = Bn
syn-1b
syn-1c
R = Me, Bn
N
H
H
OR
O
OR
O
O
O
Bn
N
O
ref. 1
Bn
2
R = Me 5 68%
6 55%
SnCl4, MeCN,
–30 °C to r.t., 6 h,
50%
N
R = Bn
TMSE = (2-trimethylsilyl)ethyl
Bn
O
O
R = TMSE anti-1a
R = Me anti-1b
Scheme 2
Lewis acid promoted rearrangements of 1,2-oxazines
R = Me
H
H
syn-1a, syn-1b, and syn-1c leading to bicyclic product 4 and novel tri-
cyclic compounds 5 or 6.
OR
N
O
Bn
TMSE = (2-trimethylsilyl)ethyl
3
The formation of the tricyclic products 3, 5 and 6 can be
rationalized by the following mechanisms. Coordination
of the Lewis acid to the ‘outer’ dioxolane oxygen of syn-
1 or anti-1, subsequent ring opening of the ketal and in-
tramolecular attack of the resulting stabilized carbenium
ion onto the enol ether moiety of the 1,2-oxazine ring lead
to stabilized carbenium ions 7 and 8 as crucial intermedi-
ates (Scheme 3). Depending on the nature of the 4-alkoxy
group different pathways are possible. With 4-(2-trimeth-
ylsilyl)ethoxy substitution the bicyclic products 2 and 4
Scheme 1 Lewis acid promoted rearrangement of 1,2-oxazines
anti-1a and anti-1b leading to bi- or tricyclic 1,2-oxazine derivatives
2 and 3.
SYNLETT 2006, No. 20, pp 3498–3500
1
8
.1
2
.2
0
0
6
Advanced online publication: 08.12.2006
DOI: 10.1055/s-2006-956463; Art ID: G31206ST
© Georg Thieme Verlag Stuttgart · New York

LETTER
Rearrangements of 1,3-Dioxolanyl-Substituted 1,2-Oxazines
3499
are formed, as this group allows for fast fragmentation of quent reactions were conducted. Depending on the time of
7 or 8 into the carbonyl compound, ethylene and a Me3SiX a hydrogenolysis it is possible to induce debenzylation
species.1 On the other hand, the methoxy group is not and N–O bond cleavage of 2-N-benzyl-protected 1,2-ox-
capable of undergoing a similarly fast fragmentation and azines.7 Hydrogenolysis of tricycles 5 and 6 first led to de-
therefore alternative steps have to follow. In the case of benzylation at the nitrogen with subsequent ring opening
carbenium ion 8 derived from anti-1b a direct cyclization of the N,O-acetal moiety affording novel 1,2-oxazines 10
under expulsion of the Lewis acid is geometrically possi- and 11 with annulated furan rings in moderate to good
ble and hence the observed tricyclic product 3 is formed. yields (Scheme 4). Acetylation of 10 provided the corre-
This pathway is not accessible in the syn-series for steric sponding 1,2-oxazine 18 which afforded suitable crystals
reasons. Instead, intermediate 7 suffers a 1,2-shift of an for an X-ray crystallographic analysis.3 This allowed not
alkyl group5 with retention of configuration resulting in only unambiguous proof of the constitution and configu-
new carbenium ion 9 which is now stabilized by the nitro- ration of 18, but also that of the structure of its precursor
gen rather than the oxygen atom. The significantly higher 5 thus supporting our mechanistic rationale as illustrated
stability of the iminium-ion-type intermediate and possi- in Scheme 3.
bly a steric relaxation are the driving forces for the rear-
rangement. Carbenium ion 9 is now able to smoothly react
with the remaining oxygen under displacement of the
Lewis acid, delivering compound 5 with a methoxy group
Hydrogenolysis of 5 for a longer period in the presence of
Boc-anhydride led to a 7:1 mixture of two constitutional
isomers, the N-protected N,O-acetals 14 and 15 in good
yield (Scheme 4). A similar result was obtained with tri-
at the central carbon atom of the tricyclic skeleton. The
cyclic precursor 6, however, now the O-benzyl group was
formation of compound 6 also follows this pathway, dem-
also removed, which provided the two isomers 16 and 17
onstrating that the 4-benzyloxy group does not undergo
(ratio: 9:1) in moderate yield. Under these reaction condi-
fragmentation into a benzyl cation at the stage of interme-
tions intermediate formation of 10 and 11 is followed by
the cleavage of the N–O bonds to furnish dihydroxy-
diate 7.6
In order to prove the structure of rearrangement products imines 12 and 13, respectively. These intermediates cy-
5 and 6 and to demonstrate their synthetic potential subse- clize and, depending on which of the two primary
hydroxyl groups is adding to the imine, either bicyclic iso-
mers 14, 16 or 15, 17 are formed. The intermediates are
O
OR
O
OR
OLA
OLA
H
H
H
H
H
N
N
O
Bn
O
Bn
O
O
H
OR
O
H
R
OLA
H
N
H
Bn
O
O
O
R
LA
R = Me
R = Bn
5
H2, Pd/C,
Boc2O, MeOH
r.t., 18 h
O
H2, Pd/C, MeOH
r.t., 1–1.5 h
O
6
Bn
N
N
Bn
O
O
8
7*
OH
OH
R = Me
O
O
R = Me, Bn
H
R = TMSE
– LA
OR
OR
NH
H
N
O
O
O
O
O
OR
LAO
OH
OH
O
H
H
R = Me
R = Me
12
10 85%
N
Bn
OR
R = Bn
R = Bn
13
11
70%
N
N
O
O
O
Bn
Bn
9*
– LA
3
2 (syn-1a-derived)
4 (anti-1a-derived)
Ac2O, DMAP, pyridine
r.t., 8 h
R = Me
84% (from 5)
* the geminal methyl groups
are omitted for clarity
OH
H
OAc
OMe
H
H
O
O
H
O
O
O
+
O
OR
NHBoc
LA = Lewis acid
OR
NHBoc
N
H
OR
O
H
HO
O
O
N
Bn
R = Me 14
R = H
18
82% (7:1)
64% (9:1)
15
17
R = Me
R = Bn
5
16
6
Scheme 4 Hydrogenolyses of rearrangement products 5 and 6
leading to 1,2-oxazine derivatives 10 and 11 or to tetrahydrofuryl
annulated tetrahydrofurans 14, 15 and 16, 17.
Scheme 3 Proposed reaction pathways for the rearrangements
leading to compounds 2, 4, 3, 5, and 6.
Synlett 2006, No. 20, 3498–3500 © Thieme Stuttgart · New York

3500
F. Pfrengle et al.
LETTER
Analytical Data for (4aR,7aS,7bS)-7-Benzyl-7b-
methoxy-4,4-dimethylhexahydro-2H,4H-1,3,6-trioxa-7-
azacyclopenta[cd]indene.
then in situ protected to generate the NHBoc moiety. In
the case of 4-benzyloxy-substituted 1,2-oxazine 6 the ben-
zyl group was also reductively removed giving com-
pounds 16, 17 with a free tertiary hydroxyl group.
[a]D22 +4.8 (c 0.42, CHCl3). 1H NMR (500 MHz, CDCl3):
d = 1.35, 1.41 (2 s, 3 H each, Me), 2.15 (dd, J = 1.1, 5.1 Hz,
1 H, 4a-H), 3.42 (s, 3 H, OMe), 3.85 (dd, J = 4.6, 10.4 Hz, 1
H, 2-H), 3.99 (dd, J = 1.1, 12.3 Hz, 1 H, 5-H), 4.00 (dd,
J = 1.0, 10.4 Hz, 1 H, 2-H), 4.01 (d, J = 14.2 Hz, 1 H,
NCH2), 4.08 (dd, J = 5.1, 12.3 Hz, 1 H, 5-H), 4.22 (d,
J = 14.2 Hz, 1 H, NCH2), 4.49 (dd, J = 1.0, 4.6 Hz, 1 H, 2a-
H), 4.50 (s, 1 H, 7a-H), 7.24–7.40 (m, 5 H, Ph) ppm. IR
(film): n = 3055–3030 cm–1 (=C–H), 2970–2870 (C–H),
1605 (C=C). MS (EI, 80 eV, 150 °C): m/z (%) = 305 (44)
[M]+, 244 (100) [M – OCH3 – CH2O]+, 214 (3) [M – C7H7]+,
91 (57) [C7H7]+. Anal. Calcd for C17H23NO4 (305.4): C,
66.86; H, 7.59; N, 4.59. Found: C, 66.51; H, 7.27; N, 4.54.
HRMS (EI, 80 eV, 150 °C): m/z calcd for C17H23NO4:
305.16272; found: 305.16366.
In summary, a novel stereoselective Lewis acid promoted
rearrangement of 1,3-dioxolanyl-substituted 1,2-oxazines
was discovered, which led to enantiopure heterocyclic
compounds with a complex skeleton. The resulting tricy-
clic products 5 and 6 are suitable precursors for further
transformations as demonstrated by their reductive trans-
formations into compounds such as 10, 11 and 14–17.
Easily accessible intermediates such as 5,6-dihydro-4H-
1,2-oxazines 10 and 11 should be excellent starting mate-
rials for addition reactions to the C=N double bond, thus
leading to new enantiopure heterocycles. All these options
will enhance the synthetic utility of 1,2-oxazines, which
has already been demonstrated in several reports.8
(5) Review concerning 1,2-sigmatropic shifts of alkyl groups:
(a) Shubin, V. G. Top. Curr. Chem. 1984, 116-117, 267; the
rearrangement described here has some similarities to the
pinacol rearrangement. (b) For Prins–pinacol tandem
reactions, see: Lavigne, R. M. A.; Riou, M.; Girardin, M.;
Morency, L.; Barriault, L. Org. Lett. 2005, 7, 5921. (c)
Also see: Overman, L. E.; Pennington, L. D. J. Org. Chem.
2003, 68, 7143.
Acknowledgment
Support by the Deutsche Forschungsgemeinschaft, the Fonds der
Chemischen Industrie and the Schering AG is most gratefully ack-
nowledged. We thank Prof. Dr. H. Hartl and I. Brüdgam for the X-
ray crystallographic analyses.
(6) In case of a p-methoxybenzyloxy substituent fragmentation
leading to bicyclic ketone 4 was observed (Pfrengle, F.;
Reissig, H.-U. unpublished results).
(7) Helms, M. Dissertation; Freie Universität Berlin: Germany,
2005.
(8) Reviews: (a) Gilchrist, T. L. Chem. Soc. Rev. 1983, 12, 53.
(b) Tsoungas, P. G. Heterocycles 2002, 57, 915.
References and Notes
(1) Al-Harrasi, A.; Reissig, H.-U. Angew. Chem. Int. Ed. 2005,
44, 6227; Angew. Chem. 2005, 117, 6383.
(2) (a) Schade, W.; Reissig, H.-U. Synlett 1999, 632.
(b) Helms, M.; Schade, W.; Pulz, R.; Watanabe, T.; Al-
Harrasi, A.; Fisera, L.; Hlobilova, I.; Zahn, G.; Reissig, H.-
U. Eur. J. Org. Chem. 2005, 1003.
(c) Tsoungas, P. G. Heterocycles 2002, 57, 1149. For
selected recent publications of our group, see: (d) Pulz, R.;
Watanabe, T.; Schade, W.; Reissig, H.-U. Synlett 2000, 983.
(e) Pulz, R.; Al-Harrasi, A.; Reissig, H.-U. Org. Lett. 2002,
4, 2353. (f) Pulz, R.; Cicchi, S.; Brandi, A.; Reissig, H.-U.
Eur. J. Org. Chem. 2003, 1153. (g) Schmidt, E.; Reissig,
H.-U.; Zimmer, R. Synthesis 2006, 2074. Selected
contributions of other groups, see: (h) Kibayashi, C.;
Aoyagi, S. Synlett 1995, 873. (i) Denmark, S. E.;
Thorarensen, A. Chem. Rev. 1996, 96, 137. (j) Streith, J.;
Defoin, A. Synlett 1996, 189. (k) Young, I. S.; Kerr, M. A.
Angew. Chem. Int. Ed. 2003, 42, 3023; Angew. Chem. 2003,
115, 3131. (l) Cardona, F.; Goti, A. Angew. Chem. Int. Ed.
2005, 44, 7832; Angew. Chem. 2005, 117, 8042.
(3) Brüdgam, I.; Hartl, H., Institut für Chemie und Biochemie,
Freie Universität Berlin, unpublished results.
(4) Typical Procedure, Conversion of syn-1b into 5.
SnCl4 (0.78 mL) was added to a solution of syn-1b (0.65 g,
2.13 mmol) in MeCN (18 mL) at –30 °C. The mixture was
allowed to warm up to 0 °C within 3 h, then stirred for an
additional 5 h at r.t., H2O (32 mL) was added and the mixture
was extracted with CH2Cl2. The combined organic extracts
were dried (Na2SO4) and concentrated. The residue was
purified by column chromatography (silica gel, hexane–
EtOAc, 2:1) to give 5 (0.44 g, 68%) as a colorless oil.
Synlett 2006, No. 20, 3498–3500 © Thieme Stuttgart · New York