Oxidation of 2-phenylhydrazono-γ-butyrolactone: A novel ring expansion rearrangement leading to tetrahydro-1,3-oxazine-2,4-dione derivatives

Article abstract of DOI:10.1039/a607830c

2-Hydroxy-2-phenylazo-γ-butyrolactone 3a, prepared from oxidation of phenylhydrazone 1a, either decomposes to form α-keto-γ-butyrolactone 2a in 80% yield or rearranges to tetrahydro-1,3-oxazine-2,4-dione derivative 4a in 95% yield under different conditio

Full text of DOI:10.1039/a607830c

View Article Online / Journal Homepage / Table of Contents for this issue
Oxidation of 2-phenylhydrazono-g-butyrolactone: a novel ring expansion
rearrangement leading to tetrahydro-1,3-oxazine-2,4-dione derivatives
Derek H. R. Barton* and Wansheng Liu*
Department of Chemistry, Texas A&M University, College Station, Texas 77843, USA
2-Hydroxy-2-phenylazo-g-butyrolactone 3a, prepared from
oxidation of phenylhydrazone 1a, either decomposes to form
a-keto-g-butyrolactone 2a in 80% yield or rearranges to
tetrahydro-1,3-oxazine-2,4-dione derivative 4a in 95% yield
under different conditions.
The intermediate 3a was surprisingly stable in the presence of
Et3N. By 1H NMR, only 10% decomposition was observed ten
days after Et3N had been added to a CDCl3 solution of 3a at
room temp. This suggested that 3a does not readily decompose
by loss of PhNNN2 anion.10 A simple explanation is that the
formation of a carbonyl group in a five-membered ring a- to
another carbonyl group is disfavoured by the strong dipole–
dipole interaction. In contrast, in open chain compounds this
transformation takes place easily8 since the opposed dipole
moments in the product cancel each other.
The chemistry of 1,3-oxazines has been of interest since the
1950s.1 Preparative schemes for many 1,3-oxazine derivatives,
especially tetrahydro-1,3-oxazines, were patented mainly as a
result of their interesting biological activities. In addition to the
synthetic utility of 5,6-dihydro-4H-1,3-oxazines,2 another fact
which makes 1,3-oxazine chemistry important is that some
antibiotics contain the 1,3-oxazine ring.1,3,4
Tetrahydro-1,3-oxazine-2,4-diones are normally synthesized
from b-hydroxy acids by reaction with sodium cyanate,
followed by cyclization with thionyl chloride.5 Alternative
preparations utilize the reaction of oxetanes with either
isocyanates or S-alkylthioureas. Here we report an unexpected
rearrangement reaction leading to tetrahydro-1,3-oxazine-
2,4-dione derivatives.
During the synthetic studies of 3-deoxy-d-manno-2-octulo-
sonic acid (KDO), we encountered the sugar derivative
2-phenylhydrazono-g-lactone 1b.6 Many failed attempts at the
conversion of 1b to KDO precursor 2b prompted us to study
oxidation of 2-phenylhydrazono-g-butyrolactone 1a (Scheme
1).7 When 1a was oxidized with bis(acetoxy)iodobenzene in
ACOH, 2-acetoxy-2-phenylazo-g-butyrolactone 3b† was iso-
lated as a bright yellow oil in 98% yield. Oxidation with
bis(trifluoroacetoxy)iodobenzene (BTIB) in TFA afforded
2-trifluoroacetoxy-2-phenylazo-g-butyrolactone 3c, which hy-
drolysed with facility when exposed to air at room temperature.
When the oxidation of 1a by BTIB was carried out in MeCN–
MeOH (5:1), crystalline 2-methoxy-2-phenylazo-g-butyro-
lactone 3d was obtained in 63% yield.
When 1a was oxidized by BTIB in MeCN–MeOH (5:1),
aqueous work-up with sodium hydrogen carbonate furnished
2-hydroxy-2-phenylazo-g-butyrolactone 3a in 80% yield after
column chromatography. Ozonolysis of 1a in methanol at
250 °C for 30 min followed by Me2S work-up also afforded 3a
in 82% yield. This intermediate behaved differently from the
acyclic counterparts.8 At room temp. 3a decomposed to form
two products, i.e. the expected 2-keto-g-butyrolactone (so-
called a-tetronic acid9) in the 2-enol tautomeric form 2a by 1H
and 13C NMR spectroscopy, and an unexpected rearrangement
product tetrahydro-1,3-oxazine-2,4-dione derivative 4a.
In CDCl3, 3a decomposed very rapidly to 2a and 4a when
catalysed by TFA. The decomposition was much slower in both
Me2CO and MeOH than in CHCl3. The formation of phenyl-
diazene or the protonated congeners was readily observed by 1H
NMR. When 1,4-benzoquinone, which is known to aid the
decomposition by electron transfer of phenyldiazene,11 was
used with TFA, 2a became the major product.
The above results suggested that a high concentration of TFA
favoured rearrangement, whereas 1,4-benzoquinone promoted
fragmentation. We propose that two different protonated forms
of 3a, i.e. 3a1 and 3a2, resulted in molecular rearrangement and
elimination of phenyldiazene, respectively (Scheme 2). Re-
NNHPh
OH
R
R
O
O
O
O
1a R = H
2a R = H
OH
OH
b R =
b R =
HO
HO
OH OH
O
OH OH
R
N
Et
Ph
OR
N
N
NPh
2
Ph
N
HO
OEt
N
1
O
O
O
O
O
O
5
3a R = H
b R = Ac
c R = C(O)CF3
d R = Me
4a R = H
b R = Ac
c R = Bz
d R = NO
e R = Et
Et2O•BF3
H
H
H
H
+
•
•
N
Ph
•
O
O
NPh
O
NPh
+
•
•
•
N
H+ transfer
(TFA used)
N
N
H
O
O
–
PhN2+Cl
PhI(O2CCF3)2
or O3–Me2S
CH2=CHOEt
TFA–CHCl3
O
O
O
O
O
NaOH–H2O
3a
(BF3 used)
3a1
3a2
1a
3a
2a
Me
92%
80–82%
80%
O
O
–
ET3O+BF4
ET2O•BF3
CH2Cl2
95%
O
O
H
CH2Cl2
83%
OH
N
Ph
N
+ PhN=NH
4a + 4e + 5
18 : 55 : 10
4a
O
O
O
4a
2a
Scheme 1
Scheme 2
Chem. Commun., 1997
571

View Article Online
moval of phenyldiazene by 1,4-benzoquinone drove the equilib-
rium between 3a1 and 3a2 towards the formation of 2a. It is
known that phenyldiazene reacts with 1,4-benzoquinone in a
stoichiometry of approximately 2:1.12 This is supported by the
observation that half an equiv. of 1,4-benzoquinone produced a
higher 2a:4a ratio. Apparently, 1,4-benzoquinone reduced the
formation of benzene. In addition to quinones, ethyl vinyl ether
in the presence of TFA was found to be highly effective in
favouring the formation of 2a. It was assumed that the driving
force came from the trapping of phenyldiazene by the
carbocation EtO+Me, which is generated by the reaction of ethyl
vinyl ether and TFA. A preparative experiment gave crystalline
2a in 80% yield when 3a was treated with TFA in the presence
of excess ethyl vinyl ether in CHCl2. Considering the
difficulties encountered by others in attempts to convert
2-oximino-g-butyrolactone to 2a,9 we conclude that the present
procedure from the phenylhydrazone provides an attractive
alternative. In principle it constitutes an efficient method of
preparing 2-keto-g-butyrolactones from simple, commercially
available starting materials.13
CO2(CH2)2CO2Me] in 61% yield. The product was identical by
1
IR, H, and 13C NMR to an authentic sample prepared from
b-propiolactone.16 Similarly, methanolysis of 4b in the pres-
ence of potassium carbonate gave N-acetylphenylhydrazine
[PhN(Ac)NH217] in 62% yield.
In conclusion, a novel rearrangement reaction provided
efficient entry to interesting N-substituted tetrahydro-1,3-ox-
azine-2,4-dione derivatives. It is conceivable that the reaction
may be extended to the construction of other related hetero-
cyclic compounds.
Financial support from the National Institutes of Health and
the Schering-Plough Corporation is gratefully acknowledged.
Footnote
† All new compounds gave satisfactory 1H and 13C NMR spectra, and
microanalytical data.
References
Interestingly, 3a readily rearranged to crystalline 4a in 95%
yield when treated with BF3·Et2O in CH2Cl2 at room temp. To
our knowledge, this seems to be the first example of a
1,2-migration of an oxycarbonyl group from carbon to azo-
nitrogen. Similarly, when 3a was treated with Meerwein’s salt
(Et3O+BF42),14 three products were isolated: 4a (18%), NA-
ethylated derivative 4e (55%), and NA,O-bis-ethylated deriva-
tive 5 (10%). When the reaction between 3a and BF3·Et2O was
carried out in CDCl3 and followed by NMR, it was observed
that an intermediate formed immediately when BF3 was added,
and then gradually decomposed to 4a at room temp. The
proposed mechanism is illustrated in Scheme 2.
Compound 4a was readily acylated and nitrosated at the NA-
position to give crystalline derivatives. Although the reaction
between 4a and Ac2O–pyridine–DMAP was very sluggish even
at elevated temperatures, the NA-acetyl derivative 4b was readily
obtained in quantitative yield at room temp. when 4a reacted
with neat acetyl chloride. The benzoyl derivative 4c was also
obtained quantitatively when benzoyl chloride was used at
60 °C. When 4a was treated with isoamyl nitrite in CH2Cl2 at
room temp., NA-nitroso compound 4d was obtained quantita-
tively after 10 min.
1 Z. Eckstein and T. Urbanski, Adv. Heterocycl. Chem., 1963, 2, 311;
Z. Eckstein and T. Urbanski, Adv. Heterocycl. Chem., 1978, 23, 1; T.
Kato, N. Katagiri and Y. Yamamoto, Heterocycles, 1980, 14, 1333.
2 A. I. Meyers, Heterocycles in Organic Synthesis, Wiley, New York,
1974, pp. 160–266; R. R. Schmidt, Synthesis, 1972, 333.
3 T. Haneishi, T. Okazaki, T. Hata, C. Tamuta, M. Namura, A. Naito,
I. Seki and M. Arai, J. Antibiot., 1971, 24, 797; K. Sasaki, Y. Kusakabe
and S. Esumi, J. Antibiot., 1972, 25, 151; Y. Kusakabe, J. Nagatsu,
M. Shibuya, O. Kawaguchi, C. Hirose and S. Shirato, J. Antibiot., 1972,
25, 44; D. B. Silvano and W. Manfred, J. Org. Chem., 1972, 42, 109.
4 M. S. von Wittenau and H. Els, J. Am. Chem. Soc., 1961, 83, 4678.
5 M. Sainsbury, in Comprehensive Heterocyclic Chemistry, ed. A. R.
Katritzky and C. W. Rees, Pergamon, New York, 1984, vol. 3, pp. 995–
1038.
6 D. H. R. Barton, J. C. Jaszberenyi, W. Liu and T. Shinada, Tetrahedron,
1996, 52, 2717.
7 R. H. Harradence and F. Lions, J. Proc. Royal Soc. N. S. Wales, 1938,
72, 221.
8 D. H. R. Barton, J. Cs. Jaszberenyi, W. Liu and T. Shinada,
Tetrahedron, 1996, 52, 14 673.
9 L. Kletz and A. Lapworth, J. Chem. Soc., 1915, 107, 1254.
10 P. C. Huang and E. M. Kosower, J. Am. Chem. Soc., 1968, 90, 2354.
11 E. M. Kosower, Acc. Chem. Res., 1971, 4, 193.
12 P. C. Huang and E. M. Kosower, J. Am. Chem. Soc., 1968, 90, 2367.
13 H. Schinz and M. Hinder, Helv. Chim. Acta, 1947, 30, 1349;
C. G. Wermuth, Bull. Soc. Chim. Fr., 1966, 1435; H. H. Wasserman and
J. L. Ives, J. Org. Chem., 1978, 43, 3238; I. Tapia, V. Alcazar,
J. R. Moran, C. Caballero and M. Grande, Chem. Lett., 1990, 697.
14 H. Meerwein, Org. Synth., 1973, Coll. Vol. 5, 1080.
15 S. Ozaki and T. Kato, J. Polym. Sci., C, 1968, 695.
16 T. L. Gresham, J. E. Jansen, F. W. Shaver, J. T. Gregory and
W. L. Beears, J. Am. Chem. Soc., 1948, 70, 1004.
17 H. Yamamoto, J. Org. Chem., 1967, 32, 3693.
Reduction of the NA-nitroso derivative 4d by zinc in AcOH
gave 4a and unsubstituted tetrahydro-1,3-oxazine-2,4-dione15
in a 1:1 ratio. The NMR and GC–MS spectra of the latter were
in agreement with the literature data. This supported the
proposed structure of 4a. When 4d was treated with zinc in
AcHO–pyridine at 0 °C, 4a was obtained in 66% yield after 30
min.
The structure of 4a was further verified through degradation
reactions. Hydrolysis of 4a by NaOH in MeOH furnished
methyl
b-(hydroxypropionyloxy)propionate
[HO(CH2)2-
Received, 15th November 1996; Com. 6/07830C
572
Chem. Commun., 1997