A mild method for the efficient [3,3]-sigmatropic rearrangement of N,O-diacylhydroxylamines

Article abstract of DOI:10.1021/jo901717a

(Chemical Equation Presented) A mild, general method for the [3,3]-sigmatropic rearrangement of N,O-diacylhydroxylamines, employing a combination of mild base and Lewis acid, is described. Employing stoichiometric amounts of reagents with respect to subst

Full text of DOI:10.1021/jo901717a

pubs.acs.org/joc
SCHEME 1. Endo’s [3,3]-Sigmatropic Rearrangement of
A Mild Method for the Efficient [3,3]-Sigmatropic
Rearrangement of N,O-Diacylhydroxylamines
N,O-Diacylhydroxylamines3
Helen Rachel Lagiakos,† Marie-Isabel Aguilar,‡ and
Patrick Perlmutter*,†
Monash University, Clayton,
Victoria, 3800, Australia, and Department of Biochemistry &
Molecular Biology, Monash University, Clayton, Victoria,
3800, Australia. †School of Chemistry. ‡Department of
Biochemistry & Molecular Biology
SCHEME 2. Alternative [3,3]-Sigmatropic Rearrangement
of N,O-Diacylhydroxylamines5
patrick.perlmutter@sci.monash.edu.au
Received August 7, 2009
enolization of 3 generates the corresponding enol and/or
enolate which then rearranges to give R-acyloxy products 4.
These products can easily be hydrolyzed to give synthetically
useful 2-hydroxyamides and acids.6 Since that report, Clark
has published a series of studies, exploring the use of
different bases in order to improve the generality of the
process.7 Two main issues with regard to substrate structure
were noted by this group. The first involved the acidity of the
R protons required to undergo enolization. It was found that
only activated systems (i.e., allylic and benzylic) would
undergo rearrangement.
In order to overcome this problem, a catalytic method
based on the use of highly basic phosphazenes in refluxing
toluene for extended periods of time was developed. Under
these conditions, even unactivated systems rearranged.
However, with such unactivated substrates, the steric nature
of the N-alkyl substituent played a significant role in pro-
moting or hindering the rearrangement. Thus, it was ob-
served that as the substituent decreased in steric demand so
did the yield. Whereas N-tert-butyl-containing substrates
gave good yields after rearrangement, N-sec-butyl and
N-isopropyl gave modest yields and N-methyl failed to react
at all.
We were interested in establishing whether milder condi-
tions than those previously described could be developed for
this class of rearrangements. We did this with a view toward
expanding the scope of the rearrangement to include as wide
a range of substrates as possible.
The rearrangement of N,O-diacylhydroxylamines em-
ploying varying ratios of a mild base, Et3N, and TMSOTf
as silylating agent was examined first (Scheme 3). These
conditions are similar to those reported by Miller’s group
for the rearrangement of N,N0-diacylhydrazides8 and
A mild, general method for the [3,3]-sigmatropic rear-
rangement of N,O-diacylhydroxylamines, employing a
combination of mild base and Lewis acid, is described.
Employing stoichiometric amounts of reagents with
respect to substrate provides R-acyloxyamides, whereas
an excess of reagents favors formation of cyclic ortho-
amides.
The [3,3]-sigmatropic rearrangement1,2 of N,O-diacylhy-
droxylamines (e.g., 1, Scheme 1) was first reported by Endo
in 1994.3 Under their reported conditions, the dienolate was
generated at low temperature which then rearranged to give
products 2 in modest to good yields.
Several years later, Clark’s group reported that under
forcing thermal conditions4 or employing an organic base,5
an alternative rearrangement can occur as shown in
Scheme 2. Thus, under Clark’s conditions, selective amide
(1) For general reviews on [3,3]-sigmatropic rearrangements: (a) Enders,
D; Knopp, M; Schiffers, R. Tetrahedron: Asymmetry 1996, 7 (7), 1847–1882.
(b) Nowicki, J. Molecules 2000, 5, 1033–1050. (c) Allin, S. M.; Baird, R. D.
Curr. Org. Chem. 2001, 5, 395–415. (d) Nubbemeyer, U. Synthesis 2003, 7,
961–1008.
(2) For general reviews on hetero-[3,3]-sigmatropic rearrangements, see:
(a) Blechert, S. Synthesis 1989, 71–82. (b) Ziegler, F. E. Chem. Rev. 1988, 88,
1423–1452. (c) Overman, L. E. Angew. Chem., Int. Ed. Engl. 1984, 23, 579–
586.
(3) (a) Endo, Y.; Uchida, T.; Hizatate, S.; Shudo, K. Synthesis 1994,
1096–1105. (b) Uchida, T.; Endo, Y.; Hizatate, S.; Shudo, K. Chem. Pharm.
Bull. 1994, 42 (2), 419–421. For Endo’s previous anionic [3,3]-sigmatropic
rearrangements of either O- or N-acylhydroxylamines, see: (c) Endo, Y.;
Hizatate, S.; Shudo, K. Tetrahedron Lett. 1991, 32 (24), 2803–2806. (d) Endo,
Y.; Hizatate, S.; Shudo, K. Synlett 1991, 9, 649–650.
(6) There are few synthetic methods available for the synthesis of 2-
hydroxyamides; see: (a) Seebach, D; Scheiss, M. Helv. Chim. Acta 1983, 66,
1618–1623. (b) Hoffman, R. V.; Nayyar, N. K.; Chen, W. J. Org. Chem. 1992,
57, 5700–5707.
(7) Clark, A. J.; Al-Faiyaz, Y. S. S.; Patel, D.; Broadhurst, M. J.
Tetrahedron Lett. 2001, 42, 2007–2009.
(8) Miller, S. J.; Bayne, C. D. J. Org. Chem. 1997, 62, 5680–5681.
(4) Al-Faiyaz, Y. S. S.; Clark, A. J.; Filik, R. P.; Peacock, J. L.; Thomas,
G. H. Tetrahedron Lett. 1998, 39, 1269–1272.
(5) Clark, A. J.; Al-Faiyaz, Y. S. S.; Broadhurst, M. J.; Patel, D.; Peacock,
J. L. J. Chem. Soc., Perkin Trans. 1. 2000, 1117–1127.
DOI: 10.1021/jo901717a
r
Published on Web 09/15/2009
J. Org. Chem. 2009, 74, 8001–8003 8001
2009 American Chemical Society

Lagiakos et al.
JOCNote
TABLE 1. Results for the [3,3]-Sigmatropic Rearrangement of N,O-Diacylhydroxylamines 5a-h (See Scheme 3)
compd
R
R1
R2
TMSOTf/Et3N
6/7a
total yield (%)
5a
5a
5b
5b
5c
5c
5d
5d
5e
5e
5f
Me
Me
Me
Me
Me
Me
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
t-Bu
t-Bu
Ph
Ph
Bn
Bn
Ph
Ph
Ph
Ph
1:1
2:2
1:1
2:2
1:1
2:2
1:1
2:2
1:1
2:2
1:1
2:2
1:1
2:2
1:1
2:2
1:0
1:10
10:1
0:1
1:0
1:0
1:0
1:4
1.2:1
1:1.5
1:0
75
90
90
85
25
60
40
90
70
75
10
50
45
70
5d
(CH2)2CHdCH2
(CH2)2CHdCH2
CH2CHdCH2
CH2CHdCH2
Ph
Ph
Me
Me
i-Pr
i-Pr
CH2CHdCH2
CH2CHdCH2
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
5f
1:0
1:0
1:0
1:0
5g
5g
5hb
5hc
2:1
25d
aProduct ratios determined by 1H NMR spectroscopy; b7 d. cReflux. dYield estimated by 1H NMR spectroscopy and crude isolation.
occurring. Also noteworthy is the successful rearrangement
of 5c, where both acyl substituents are allylic, and R = Me,
as such unactivated systems were inaccessible employing
base catalysis alone.
SCHEME 3. [3,3]-Sigmatropic Rearrangement of N,O-Diacyl-
hydroxylamines 5a-h with Et3N and TMSOTf
Compounds 5d-g, all bearing the more synthetically
useful N-benzyl substituent, were prepared and rearranged
in order to establish the generality of this new method.
As evident from Table 1, our conditions were able to
rearrange both activated and nonactivated systems. In line
with the literature,7 branched substituent 5f reacted slug-
gishly, though with excess reagent, a 50% yield was obtained.
Substrate 5h proved the most difficult to rearrange, requiring
a 7 days at room temperature, or overnight reflux, to pro-
vide, at best, only poor yields (25%) of impure material.
Hence treating N,O-diacylhydroxylamines with 1 equiv
each of TMSOTf and base produces silyl ketene aminals,
which can then undergo [3,3]-sigmatropic rearrangement to
give the corresponding silyliminoethers. Without significant
excess of TMSOTf quenching of the reaction yields 6.
However, where an excess of TMSOTf is present a second
product, 7, is generated and, in some cases, becomes the
major product (Scheme 3). Although the precise mechansim
of formation of this second product has not been studied,
it is noteworthy that Kamimura9,10 has demonstrated that
R-hydroxyamides are readily converted into cyclic ortho-
amides under conditions similar to those employed in this
study.
Kamimura’s group for the formation of cyclic orthoamides
from R-hydroxyamides.9,10
Thus, 5a was first treated with 1 equiv each of Lewis acid
and base at -78 °C and then stirred overnight at room
temperature. The rearranged product 6a was obtained in
75% yield. When a 1-fold excess of base and silylating agent
was employed, a 10:1 ratio of 7a/6a was obtained in 90%
yield. With these encouraging results in hand, we examined
variations in substituents R, R1, and R2 as well as ratios of
reagents to substrate. This correlation between major pro-
ducts and reagent to substrate ratios was found to hold in
most cases. The results are collected in Table 1.
Finally, we briefly evaluated the use of a chiral O-acyl
moiety to induce diastereoselectivity during the rearrange-
ment (Scheme 4). The reaction proceeded in modest yield but
provided essentially an equal mixture of the two diastereo-
mers. Separation of the two compounds was easily achieved
with column chromatography.
Assignment of the stereochemistry of each diastereomeric
product could be achieved by acylation of mandelamide.
Thus, reaction of N-benzyl-(S)-mandelamide (11) with
(S)-(O-methyl)mandelic acid (12) gave a product which
was identical in every way with 10, generated by rearrange-
ment of 8 (Scheme 5). The same process was repeated to
confirm diastereomer 9.
Compound 5b followed the trend established by the initial
experiment, giving products 6b or 7b11 in excellent yields
depending on the number of equivalents used. The rearran-
gement was also attempted at lower temperature but resulted
in no reaction. For those cases where both the ester and
amide moieties bear acidic R-protons, i.e., 5b and 5c, only
the amide moiety undergoes enolization under these con-
ditions.9,10 This observation is important as it precludes
the possibility of the alternative, “Endo”, rearrangement3
(9) Kamimura, A.; Omata, Y.; Kakehi, A.; Shirai, M. Tetrahedron 2002,
58, 8763–8770.
(10) (a) Omata, Y; Kakehi, A; Masashi, S; Kamimura, A. Tetrahedron
Lett. 2002, 43, 6911–6914. (b) Kamimura, A.; Omata, Y.; Keiichi, T.; Shirai,
M. Tetrahedron 2003, 59, 6291–6299. (c) Kamimura, A.; Tanaka, K.;
Hayashi, T.; Omata, Y. Tetrahedron Lett. 2006, 3625–3627.
(11) Most of the cyclic orthoamides were stable in solution. However, 7b
proved to be the exception and decomposed when left overnight in a CDCl3
solution.
In conclusion we have developed a new, reliable protocol for
the [3,3]-sigmatropic rearrangement of N,O-diacylhydroxyla-
mines. This pathway allows for milder reaction conditions and
a wider substrate range than previously possible.
8002 J. Org. Chem. Vol. 74, No. 20, 2009

Lagiakos et al.
JOCNote
solid which upon spectroscopic analysis, proved to be pure 6a
(150 mg, 75%): mp 142-143 °C; 1H NMR (300 MHz, CDCl3)
δ 8.10-8.14 (m, 2H), 7.28-7.65 (m, 8H), 6.37 (s, 1H), 6.20 (brs,
1H), 2.90 (d, J=4.8, 3H); 13C NMR (75 MHz, CDCl3) 169.1,
165.1, 135.8, 133.8, 130.0, 129.5, 128.9, 128.8, 128.8, 127.5, 76.1,
26.4; FT-IR νmax/cm-1 3311, 3092, 3065, 3036, 2949, 1725, 1661,
1561, 1496, 1449, 1315, 1245, 1113, 1025, 985, 708, 683; ESI-
HRMS m/z 270.1124, 292.0947 (C16H15NO3 þ Hþ requires
270.1130, C16H15NO3 þ Naþ requires 292.0950).
SCHEME 4. [3,3]-Sigmatropic Rearrangements of N,O-Dia-
cylhydroxylamine 8 Bearing a Homochiral O-Acyl Group
Preparation of 3-Methyl-2,5-diphenyl-2-(trimethylsilyloxy)-
oxazolidin-4-one (7a). N-(Benzoyloxy)-N-methyl-2-phenylace-
tamide (5a) (100 mg, 0.371 mmol) was dissolved in 1.5 mL of
dry DCM and the temperature lowered to -78 °C. To this was
added TMSOTf (135 μL, 0.743 mmol), followed 5 min later by
NEt3 (104 μL, 0.743 mmol). After being stirred at this tempera-
ture for 30 min, the reaction was removed from the cooling bath,
allowed to warm to room temperature, sealed, and left to stir
overnight. The reaction was quenched by pouring directly onto
a silica gel plug and eluted with EtOAc. Solvent removal in
vacuo left a white solid which upon spectroscopic analysis
showed a 10:1 mixture of 7a/6a in a combined 90% yield.
Column chromatography (20% EtOAc/hexane) afforded 7a as
a white solid (102 mg, 80%): mp 62-63 °C; 1H NMR (400 MHz,
CDCl3) δ 7.20-7.62 (m, 10H), 5.52 (s, 1H), 2.70 (s, 3H), 0.22
(s, 9H); 13C NMR (100 MHz, CDCl3) 169.5, 139.5, 135.5, 129.5,
128.9, 128.7, 128.7, 128.4, 127.0, 126.9, 126.8, 110.7, 78.6, 25.7,
1.2; FT-IR νmax/cm-1 3035, 2957, 1716, 1424, 1231, 1090, 1062,
1035, 869, 839, 760, 698; ESI-HRMS m/z 342.1519 (C19H23-
NO3Si þ Hþ requires 342.1525).
SCHEME 5. Chemical Proof of the Stereochemistry of Rear-
rangement Product 10
Experimental Section
Typical Procedures for the [3,3]-Sigmatropic Rearrangement
of N,O-Diacylhydroxylamines under Conditions Which Provide
Either Product 6 or 7. Preparation of 2-(Methylamino)-2-oxo-
1-phenylethyl Benzoate (6a). Under an argon atmosphere,
N-(benzoyloxy)-N-methyl-2-phenylacetamide (5a) (200 mg,
0.743 mmol) was dissolved in 3 mL of dry DCM and the
temperature lowered to -78 °C. To this, TMSOTf (135 μL,
0.743 mmol) was added, followed 5 min later by NEt3 (104 μL,
0.743 mmol). After being stirred at this temperature for 30 min,
the reaction was removed from the cooling bath, allowed to
warm to room temperature, sealed and left to stir overnight. The
reaction was quenched by pouring directly onto a silica gel plug
and eluted with EtOAc. Solvent removal in vacuo left a white
Acknowledgment. Support for this work through the
award of an Australian Research Council Discovery Grant
(No. DP0557486) is gratefully acknowledged.
Supporting Information Available: Experimental proce-
1
dures, characterization data, and copies of H and 13C NMR
spectra for all new compounds. This material is available free of
charge via the Internet at http://pubs.acs.org.
J. Org. Chem. Vol. 74, No. 20, 2009 8003