Asymmetric [3,3]- and [1,3]-Sigmatropic rearrangements of γ-allyloxy vinylogous urethanes

Article abstract of DOI:10.1021/ol901679h

Vinylogous urethanes derived from condensation of prolinol or prolinol ferf-butyldimethylsilyl ether with 4-allyloxyketoester were found to undergo a thermal [3,3]-sigmatropic rearrangement, providing compounds with N-substituted quaternary carbon centers

Full text of DOI:10.1021/ol901679h

ORGANIC
LETTERS
2009
Vol. 11, No. 18
4224-4227
Asymmetric [3,3]- and [1,3]-Sigmatropic
Rearrangements of γ-Allyloxy
Vinylogous Urethanes
Yu-Jang Li,*,† Yuan-Kang Chang,‡ Guo-Ming Ho,† and Hua-Ming Huang†
Department of Applied Chemistry, National Chiayi UniVersity, 300 UniVersity Road,
Chiayi City, 600 Taiwan, and Department of Applied Chemistry, Chaoyang UniVersity
of Technology, 168 Gifeng East Road, Wufeng, Taichung County, 416 Taiwan
yjli@mail.ncyu.edu.tw
Received July 22, 2009
ABSTRACT
Vinylogous urethanes derived from condensation of prolinol or prolinol tert-butyldimethylsilyl ether with 4-allyloxyketoester were found to
undergo a thermal [3,3]-sigmatropic rearrangement, providing compounds with N-substituted quaternary carbon centers. Cyclizations
(subsequently or in situ) of the rearranged products generated hexahydro-3,4-dioxa-8a-aza-as-indacen-2-ones. Various terminally substituted
allyloxy ketoesters and arylmethoxy ketoesters were found to generate tricyclic compounds via [1,3]-sigmatropic rearrangement. Finally, tricyclic
lactones were transformed successfully into lactams.
During the course of our investigation into [2,3]-Wittig
rearrangements utilizing γ-allyloxy vinylogous urethanes 3
as starting material, trace amounts of aldehyde 4 were
detected during the condensation of pyrrolidine with ke-
toesters 2.1 The aldehyde was presumably generated via a
[3,3]-sigmatropic rearrangement pathway (the Claisen rear-
rangement) between the isomerized double bond and the
allyloxyl moiety.2 [3,3]-Sigmatropic rearrangement of a
similar system 5, generated by catalytic aminomercuriation
of propagyl allyl ether, was previously reported by Barlu-
enga.3 More recent rearrangement studies of 6 and 7 reported,
respectively, by Kazmaier and Hruby led to the syntheses
of syn- and anti-ꢀ-substituted γ,δ-unsaturated amino acids
(Figure 1).4,5
We envisioned that the rearrangement of the system 3
would constitute a general protocol for the construction of a
chiral N-substituted quaternary carbon center if a chiral
pyrrolidine was involved.6 Therefore, syntheses of chiral
(3) Barluenga, J.; Aznar, F.; Liz, R.; Bayod, M. J. Org. Chem. 1987,
52, 5190.
† National Chiayi University.
‡ CYUT.
(4) (a) Kazmaier, U. Angew. Chem., Int. Ed. 1994, 33, 998. (b) Kazmaier,
U.; Krebs, A. Angew. Chem., Int. Ed. 1995, 34, 2012. (c) Kazmaier, U.;
Maier, S. J. Org. Chem. 1999, 64, 4574. (d) Kazmaier, U.; Mues, H.; Krebs,
(1) Li, Y.-J.; Lee, P.-T.; Yang, C.-M.; Chang, Y.-K.; Weng, Y. -C.;
Liu, Y.-H. Tetrahedron Lett. 2004, 45, 1865.
(2) For reviews of the Claisen rearrangement, see: (a) Ziegler, F. E.
Chem. ReV 1988, 88, 1423. (b) Ito, H.; Taguchi, T. Chem. Soc. ReV. 1999,
28, 43. (c) Hiersemann, M.; Abraham, L. Eur. J. Org. Chem. 2002, 1461.
(d) Nubbemeyer, U. Synthesis 2003, 961.
A. Chem. Eur. J. 2002, 8, 1850.
(5) (a) Qiu, W.; Gu, X.; Soloshonok, V. A.; Carducci, M. D.; Hruby,
V. J. Tetrahedron Lett. 2001, 42, 145. (b) Qu, H.; Gu, X.; Min, B. J.; Liu,
Z.; Hurby, V. J. Org. Lett. 2006, 8, 4215
.
10.1021/ol901679h CCC: $40.75
Published on Web 08/25/2009
 2009 American Chemical Society

Table 1. Rearrangement of Various Alkene Substrates
Figure 1. 3 and other known enol ether enamine systems.
vinylogous urethanes incorporated with chiral pyrrolidine
were commenced to investigate this rearrangement.
Compounds 10a,b were first synthesized by condensa-
tion of ketoesters 8a,b with prolinol tert-butyldimethylsilyl
ether 9 in 98% and 95% yield, respectively. When 10a,b
were heated, using toluene as solvent, they provided
inseparable diastereomeric aldehydes 11a,b. While 10a,b
were stable enough to be isolated, attempts to isolate 10c
under the same conditions gave mixtures of 10c and 11c.
It is conceivable that 11c was generated due to the
prolonged heating in the synthesis of 10c. Therefore, 11c
was obtained simply by heating of 8c with prolinol tert-
butyldimethylsilyl ether 9 in toluene. Subsequent removal
of the O-silyl protecting group of 11a-c by tetra n-
butylammonium fluoride (TBAF) resulted in the formation
of syn- and anti-hexahydro-3,4-dioxa-8a-aza-as-indacen-
2-ones 12a-c in 87-89% yields (Table 1). The anti
stereochemistry of the minor product 12a-anti was con-
firmed by X-ray analysis (Figure 2).7
a Mixtures of inseparable diastereomers. b Isolated yields.
The success of this rearrangement was found to be highly
influenced by the substitution on the allyl group. Having
substituents on R3, R4, or R5 generally led to the formation
of polymerized products that would most likely have resulted
from the decomposition of aldehydes.
methyl group seems to insert a synergistic effect with the
chiral arm, therefore providing exclusively 14 in 90% yield.
When terminally substituted propenoxy ketoesters 15a-c
were used in the rearrangement studies (Table 3), 15a (R4
) Pr) provided [3,3]-rearranged product 16-syn in 65% yield.
Although absolute stereochemistries of the resulting propyl
To avoid the decomposition problem, prolinol was used
as a surrogate to react with various 4-allyloxy substituted
ketoesters. It is conceivable that the intramolecular trapping
of the resulting aldehyde by the hydroxymethyl moiety to
form hemiacetal together with the following cyclization
to form a lactone would both efficiently protect the aldehyde
from decomposition and provide thermodynamic stability for
the forward rearrangement. Therefore, when L-prolinol was
used to react with terminal unsubstituted allyloxy ketoesters
(R4 ) R5 ) H), it provided tricyclic compounds in better
yields than in prolinol tert-butyldimethylsilyl ether cases
(Table 2). In entries 1 and 2, the rearrangement provided
reasonable diastereoselectivities, whereas in entry 3 with
cumbersome gem-dimethyl substitution the rearrangement
succumbed to 2 to 1 selectivity. In entry 4, the internal vinylic
1
group were not yet determined, 300 MHz H NMR in
D-chloroform indicated that the ratios of diastereomers were
approximately 2/1. When bulkier substituted 15b (R4 ) Ph)
and 15c (R4 ) R5 ) Me) were used in the rearrangement
studies, 15b provided 17a-syn and 17b-syn/anti in 73% yield
and in 5:75:20 ratio. In comparison to the propyl substitution
on 16-syn, phenyl substitution on 17a-syn turned out to be
(6) For recent reviews regarding synthesis of N-substituted quaternary
carbon, see :(a) Kang, S. H.; Kang, S. Y.; Lee, H.-S.; Buglass, A. J. Chem.
ReV. 2005, 105, 4537. (b) Ohfune, Y.; Shinada, T. Eur. J. Org. Chem. 2005,
5127. (c) Cativiela, C.; D´ıaz-de-Villegas, M. D. Tetrahedron: Asymmetry
2007, 18, 569.
(7) CCDC 678877 & CCDC 678876 contain the supplementary crystal-
lographic data for this paper. These data can be obtained free of charge
from the Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
Figure 2. X-ray crystallographic structure (ellipsoid: 50% prob-
ability) of 12a-anti.
Org. Lett., Vol. 11, No. 18, 2009
4225

Table 2. Rearrangement of Various Alkene Substrates
Table 4. Relative Product Composition of [3,3]- vs
[1,3]-Sigmatropic Rearrangement under Different Solventsa
solvent
bp (°C)
19 (%)
18-syn (%)
12c-syn/anti (%)
(syn/anti)
entry
substrate
productsa (yield)
ratiob
benzene
toluene
xylene
DMF
a
80
110
144
153
52
10
28
30
10
20
60
90
1
2
3
4
R1 ) R2 ) R3 ) H, 8a
R1 ) Me, R2 ) R3 ) H, 8b
R1 ) R2 ) Me, R3 ) H, 8c
R3 ) Me, R1 ) R2 ) H, 13
b
84
12a (5:1)
81
12b (9:1)
100
80
90
12c (2:1)
1
14 (syn only)
Ratio determined by the integration from H NMR signals.
a Isolated yields. Integration from H NMR signals.
1
exclusively compound 21 with no [1,3]-rearrangement
(Figure 3).10
stereospecific. Nevertheless, the absolute stereochemistry
could not be determined. Ketoester 15c rearranged to 18a-
syn and 12c-syn/anti in 80% yield and in 6:51:43 ratio.
Products 17b-syn/anti and 12c-syn/anti were presumably
generated via a previously unforeseen [1,3]-sigmatropic
rearrangement pathway (Table 3).8
Temperature-dependent rearrangement of 15c under dif-
ferent solvents was studied as shown in Table 4. Although
reaction for 1 h provided about a 3:2 ratio of [3,3]- to [1,3]-
sigmatropic rearrangement products in benzene, it provided
exclusively the [1,3]-rearranged product when DMF was used
as refluxing solvent (Table 4).
Figure 3. Deuterium labeling studies of rearrangement of 20.
To account for the diastereoselectivity and the pathway
preference between [3,3]- vs [1,3]-rearrangement products,
Scheme 1 was depicted to briefly explain our observations.
To clarify whether [1,3]-sigmatropic rearrangement should
be involved in the simple allyl substituted case (R1∼R5 )
H), a deuterium-labeled compound 20 was subjected to the
study.9 Careful NMR analysis indicated that the compound
underwent pure [3,3]-sigmatropic rearrangement to generate
Scheme 1
.
Reasoning for the syn/anti Ratio and [3,3]- vs
[1,3]-Reaction Pathway
Table 3. Rearrangement of R4, R5 Substituted Alkene Substrates
The chiral arm (CH2OH) on the pyrrolidine was expected
to hamper the approach of the allyloxy moiety from the
(8) Recent review of [1,3] O-to-C rearrangement, see: Nasveschuk, C. G.;
Rovis, T. Org. Biomol. Chem. 2008, 6, 240.
(9) Deuterated compound 20 was obtained via the reduction of acrolein
with sodium borodeuteride, follow by the reaction with methyl 4-bromoac-
etoacetate.
b
a Isolated yields. Integration from H NMR signals.
1
4226
Org. Lett., Vol. 11, No. 18, 2009

R-face in the TSA′ and therefore bias the reaction to proceed
via TSA in favor of generating the syn product as the major
diastereomer. Although moderate stereoselectivity of the
resulting azaquaternary center was observed, stereoselectivity
of the substitution on the neighboring allylic position ranged
from nonselective (R4 ) Pr, Table 3, entry 1) to excellent
(R4 ) Ph, Table 3, entry 2). Presumably, it depends on the
degree of preferences between TSA-chair and TSA-boat.
While the less hindered allyloxy moiety (Table 2, entries
1-4, and Table 3, entry 1) rearranged exclusively by the
[3,3]-pathway, the more hindered case (Table 3, entries 2
and 3) was observed to rearrange mainly by the [1,3]-
pathway. Presumably the steric interaction between the two
reacting terminals prohibits the [3,3]-pathway.
After completion of the investigations on allyloxy systems,
we turned our attention to the study of the substitution pattern
when allyloxy double bonds are part of aromatic systems.
Arylmethoxy-substituted ketoesters such as 1-benzyloxy,
1-thiophenomethoxy, and 1-furanomethoxy 22a-c were first
synthesized in 71-75% yields, respectively. Though the
phenyl case provided 45% of [1,3]-rearranged tricyclic
product 23, thienyl and furanyl provided moderate yields of
[1,3]-rearranged products 24 and 25a-syn/anti.11 Meanwhile,
in the furanyl case, trace amounts of [3,3]-rearranged product
25b were observed and carefully isolated (Scheme 2).
Figure 4. X-ray crystallographic structure (ellipsoid: 50% prob-
ability) of 25a-syn
.
benzylamine and trifluoroacetic acid in toluene at reflux to
generate 3-benzyl-octahydro-4-oxa-3,8a-diaza-as-indacen-2-
ones in 80-90% yields (Scheme 3).
Scheme 3. Transformation of Tricyclic Lactones into Lactams
Scheme 2. Rearrangement of Arylmethoxy-Substituted Substrate
In summary, we have reported the studies of [3,3]- and
[1,3]-sigmatropic rearrangements of the γ-allyloxy and
arylmethoxy vinylogous urethane systems. A chiral N-
substituted quaternary cabon center was successfully ob-
tained, and tricyclic lactones can be transformed successfully
into lactams. Further studies and their applications to natural
product synthesis are currently in progress in our laboratory.
Acknowledgment. We thank the Taiwan National Science
Council (NSC-94-2113-M-415-003) for generous financial
support. Database service from the NCHC and partial support
from the mass spectrometer facility provided by NCHU and
support from the X-ray facility of NTU are also acknowl-
edged.
A single crystal of 25a-syn was analyzed by X-ray
crystallography to confirm its stereochemistry as well as the
stereochemistries of all the other related syn-compounds7
(Figure 4).
Finally, tricyclic lactones could be successfully converted
to the lactams 26-30, simply by reacting them with
Supporting Information Available: Full experimental
(10) Proton-decoupled 13C spectra showed a 1:1:1 triplet at 119.9 ppm
for the terminal carbon of alkene (CH2-CHdCHD) and a pure singlet at
39.4 ppm for the methylene carbon (CH2-CHdCHD).
1
details and characterization including H and 13C spectro-
scopic data for all new compounds. This material is available
free of charge via the Internet at http://pubs.acs.org.
(11) [1,3]-Rearrangement involving the arylmethyl shift from oxygen
to carbon, see: (a) Burger, K.; Gaa, K.; Geith, K.; Schierlinger, C. Synthesis
1989, 850. (b) Shishido, K.; Shitara, E.; Fukumoto, K. J. Am. Chem. Soc.
1985, 107, 5810.
OL901679H
Org. Lett., Vol. 11, No. 18, 2009
4227