Control of electron demand in the cycloadditions of 2(H)-1,4-oxazin-2-ones

Article abstract of DOI:10.1016/j.tetlet.2004.07.105

3-Methoxy-5-chloro-6-methyl-2(H)-1,4-oxazin-2-one 4, 3-methoxy-5-chloro-6- phenyl-2(H)-1,4-oxazin-2-one 5, 3-phenylsulfenyl-5-chloro-6-methyl-2(H)-1,4- oxazin-2-one 6, and 3-phenylsulfenyl-5-chloro-6-phenyl-2(H)-1,4-oxazin-2-one 7, are ambident dienes and undergo Diels-Alder cycloadditions with electron neutral, rich and deficient dienophiles.

Full text of DOI:10.1016/j.tetlet.2004.07.105

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 7121–7124
Control of electron demand in the cycloadditions of
2(H)-1,4-oxazin-2-ones
a
b
a
Kamyar Afarinkia,a, Akmal Bahar, Judi Neuss and Maushami Vyas
*
aDepartment of Chemistry, King’s College, Strand, London WC2R 2LS, UK
bCelltech R&D Ltd., 208 Bath Road, Slough, Berkshire, SL1 3WE, UK
Received 8 June 2004; revised 14 July 2004; accepted 22 July 2004
Available online 13 August 2004
Abstract—3-Methoxy-5-chloro-6-methyl-2(H)-1,4-oxazin-2-one 4, 3-methoxy-5-chloro-6-phenyl-2(H)-1,4-oxazin-2-one 5, 3-phenyl-
sulfenyl-5-chloro-6-methyl-2(H)-1,4-oxazin-2-one 6, and 3-phenylsulfenyl-5-chloro-6-phenyl-2(H)-1,4-oxazin-2-one 7, are ambident
dienes and undergo Diels–Alder cycloadditions with electron neutral, rich and deficient dienophiles.
Ó 2004 Elsevier Ltd. All rights reserved.
5-Chloro-2(H)-1,4-oxazin-2-ones bearing an alkyl or
phenyl substituent at their 6-position and a chloro, alkyl
or phenyl substituent at their 3-position, for example,
1a–d, are ambident azadienes and undergo Diels–Alder
cycloadditions with electron neutral, rich and deficient
dienophiles (Scheme 1).1–7 These cycloadditions are
highly efficient, but with the exception of those for 1d,
are poorly regio- and stereoselective (Table 1). Although
the cycloadditions of 5-chloro-2(H)-1,4-oxazin-2-ones
have proved to be very useful synthetically,8–10 the lack
of selectivity in the reactions is a major obstacle in the
development of new methodologies based around them.
Therefore, methods which allow improved selectivity in
Scheme 1. Cycloadditions of 2(H)-1,4-oxazin-2-ones with non-polar
substituents.1–4
these cycloadditions are highly desirable.
Although the ambident nature of 2(H)-1,4-oxazin-2-
We expected that introduction of electronically biased
groups onto the 2(H)-1,4-oxazin-2-one ring should
ones is unique amongst azadienes, it is paralleled in
the chemistry of substituted 2(H)-pyran-2-ones.11–17 Py-
rone dienes bearing electronically neutral substituents
also undergo poorly selective cycloadditions with dieno-
philes. However, the efficiency, as well as the regio- and
stereoselectivity, of these cycloadditions improves signifi-
cantly if the 2(H)-pyran-2-one diene is electronically
matched with the dienophile by means of ring substitu-
tion. Hence, 3-phenylsulfenyl-2(H)-pyran-2-one 218,19
and 3-phenylsulfonyl-2(H)-pyran-2-one 320–23 react with
electron deficient and electron-rich dienophiles, respect-
ively, with excellent regio- and stereocontrol (Scheme 2).
Table 1.
R1
R2
R3
8-endo
7-endo
8-exo
7-exo
Me
Cl
Cl
CO2Me
OEt
35
73
55
60
33
0
32
27
27
40
0
0
0
0
Me
Me
CO2Me
OBu
18
0
Ph
Cl
Cl
CO2Me
OBu
30
30
20
20
30
30
20
20
0
Ph
Ph
Ph
CO2Me
4-BrC6H4
OBu
80
10
Trace
0
10
Trace
Trace
Keywords: 2(H)-1,4-Oxazin-2-ones; Diels–Alder; Cycloaddition.
*
>98
>98
0
Corresponding author. Tel.: +44-2078482422; fax: +44-
2078482810; e-mail: kamyar.afarinkia@kcl.ac.uk
0
0040-4039/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.07.105

7122
K. Afarinkia et al. / Tetrahedron Letters 45 (2004) 7121–7124
Scheme 2. Cycloadditions of 2(H)-pyran-2-ones.18–23
Scheme 4. Cycloadditions of 4.
Table 2.
R
Ratio of cycloadductsa (yield %)b
8-endo
7-endo
8-exo
7-exo
CO2Me
CN
67 (53)
50 (46)
57 (46)
54c
0
33 (26)
50 (44)
43 (34)
46c
0
Scheme 3. Preparation of 2(H)-1,4-oxazin-2-ones.
0
0
4-BrC6H4
OAc
OBu
0
0
Trace
24 (18)
Trace
Trace
50 (38)
25 (19)
similarly improve the regio- and stereoselectivity in their
cyclo- addition reactions. Therefore, we decided to
investigate the cycloaddition chemistry of 5-chloro-
2(H)-1,4-oxazin-2-ones bearing electron donating
(OMe, SPh) substituents at their 3-position. 5-Chloro-
3-methoxy-6-methyl-2(H)-1,4-oxazin-2-one 4, and 5-
a Ratios of cycloadducts are normalised.
b Yields after purification by chromatography.
c Cycloadducts decomposed during chromatography.
chloro-3-methoxy-6-phenyl-2(H)-1,4-oxazin-2-one
5,
were prepared by treatment of 1a and 1c with methanol-
ic hydrogen chloride. 5-Chloro-3-phenylsulfenyl-6-
methyl-2(H)-1,4-oxazin-2-one 6, and 5-chloro-3-phenyl-
sulfenyl-6-phenyl- 2(H)-1,4-oxazin-2-one 7, were pre-
pared by treatment of 1a and 1c with thiophenol and
aluminium trichloride (Scheme 3).4
The 2(H)-1,4-oxazin-2-ones were treated with a range of
electron rich and electron deficient dienophiles (8–13,
Fig. 1) at room temperature or at 70°C, and the crude
reaction mixtures were analysed by NMR. In all cases,
clean conversion was observed with only the cycload-
ducts and traces of starting materials present in the
crude reaction mixtures. However, when chromatogra-
phy was used to isolate the cycloadduct products, this
resulted in partial or complete decomposition of the cy-
cloadducts in some cases and loss of yield. The results
are shown in Scheme 4 and Table 2, and Scheme 5
and Table 3. Assignments of the relative configurations
of isolated cycloadducts were based on NOE (nuclear
Overhauser enhancement) data,1 and by comparison of
the chemical shifts with those of analogous com-
pounds.1,6,10
Scheme 5. Cycloadditions of 6 and 7 with mono-substituted dieno-
philes.
Table 3.
R1
R2
Ratio of cycloadductsa (yield %)b
8-endo
7-endo
8-exo
7-exo
Me
CO2Me
4-BrC6H4
OCH2CH2Cl
50 (35)
75 (69)
50 (18)c
0
50 (25)
25 (15)
25 (10)c
0
0
0
12.5c
12.5c
Ph
CO2Me
CN
32 (27)
50 (30)c
58 (57)
33d
18 (15)
16 (4)c
18 (12)
26d
44 (38)
34 (5)c
14 (3)
33d
6 (1)
0
4-BrC6H4
OAc
OCH2CH2Cl
10 (10)
8d
0
48 (33)c
17 (13)c
35 (27)c
a Ratios of cycloadducts are normalised.
b Yields after purification by silica gel chromatography.
c Cycloadducts partially decomposed during chromatography.
d Cycloadducts decomposed during chromatography.
Figure 1.

K. Afarinkia et al. / Tetrahedron Letters 45 (2004) 7121–7124
7123
5-Chloro-3-methoxy-6-methyl-2(H)-1,4-oxazin-2-one 4,
underwent highly efficient cycloadditions with dieno-
philes 8–12 (Scheme 4, Table 2). In the reactions with
electron-neutral and electron-rich dienophiles, all four
possible cycloadducts were seen in the crude NMR.
However, in the reactions with electron-deficient dieno-
philes, only the 8-endo and 8-exo cycloadducts were seen
in the crude NMR.
The results from the cycloadditions of 4 with electron-
deficient and electron-neutral dienophiles were entirely
consistent with our expectations of the electron demand
of this compound. Having an electron donating meth-
oxy substituent at its 3-position, compound 4 should
be an electron-rich diene and favour normal electron de-
mand cycloadditions with electron-deficient dienophiles.
The regioselectivity in favour of the formation of the 8-
substituted cycloadducts is expected from AlderÕs rules,
even though the stereoselectivity in favour of the endo
cycloadducts is disappointingly low. Surprisingly, we
also obtained cycloadducts with electron-rich dieno-
philes. These reactions are neither regio- nor stereoselec-
tive as might be expected from the mis-match from the
electronic demand of the diene and dienophile.
Scheme 6. Cycloadditions of 4 and 6 with disubstituted dienophiles.
Table 4.
Dienophile
Diene
Ratio of cycloadductsa (yield %)b
8-endo
7-endo
8-exo
7-exo
14
14
15
15
4
6
4
6
50 (16)
75 (20)
50 (15)
20 (13)
0
50 (5)
25 (10)
0
0
0
0
0
50 (25)
80 (38)
0
0
The results from the cycloadditions of compounds 6 and
7 with mono-substituted dienophiles mirrored those ob-
tained from the cycloadditions of compound 4 (Scheme
5, Table 3) Again, the results were consistent with our
expectations of the electron demand of these com-
pounds. Having an electron donating phenylsulfenyl
substituent at their 3-position, compounds 6 and 7 are
electron-rich dienes and favour normal electron demand
cycloadditions with electron deficient dienophiles.
Exclusive formation of 8-substituted cycloadducts in
reactions with electron-deficient dienophiles is expected
although we again obtained poor stereoselectivity as
was observed in the cycloadditions of compound 4.
Again we obtained cycloadducts with electron-rich dien-
ophiles in reactions that were neither regio- nor stereose-
lective.
a Ratios of cycloadducts are normalised.
b Yields after purification by silica gel chromatography.
NMR spectra in any of the cases. The regiochemistry
of the cycloadducts were confirmed by NOESY. For in-
stance, the regiochemistry of the cycloadducts 16a and
17a were confirmed by an NOE between the signals of
H-7 at 1.90 and 2.66ppm for 16a, and 2.05 and
2.75ppm for 17a, with the methyl substituent at C-1.
Similarly, the regiochemistry of the cycloadducts 18a
and 19a were confirmed by an NOE between the signals
of H-7 at 2.49 and 2.67ppm, respectively, with the
methyl substituent at C-1.
The observed regiochemical preference in the cycloaddi-
tions of 4 and 6 with 14 are in line with that observed in
the cycloaddditions with methyl acrylate 8. However,
the outcome of the cycloadditions of 4 and 6 with 15
are unexpected. The dominance of the 7-substituted
cycloadducts in these reactions indicates that the regio-
chemical preferences in the cycloadditions are not exclu-
sively determined by the electronic demand of the
dienes.
It is clear from these results that the 5-chloro-2(H)-1,4-
oxazin-2-ones bearing electron-donating groups at their
3-positions remain ambident dienophiles and are still
capable of reacting with electron-rich, electron-neutral
and electron-deficient dienophiles. The advantage
gained from introducing electron-donating substitu-
ents at the 3-position of 5-chloro-2(H)-1,4-oxazin-2-
ones is that cycloadditions with their electronically
matched dienophiles (i.e., electron deficient or electron
ÔneutralÕ alkenes) become regioselective.
Another unusual feature of the cycloadditions of these
unprecedented ambident dienophiles is that 5-chloro-6-
phenyl-2(H)-1,4-oxazin-2-ones, 5 and 7, are considera-
bly less reactive than the corresponding 5-chloro-6-
methyl-2(H)-1,4-oxazin-2-ones, 4 and 6. For instance,
2(H)-1,4-oxazin-2-ones, 5 and 7, do not afford any
cycloadducts with 15 although they do undergo efficient
cycloaddition with 14 to afford 8-substituted cycload-
ducts 20 and 21 (Scheme 7 and Table 5).
Cycloadditions of 2(H)-1,4-oxazin-2-ones with di-substi-
tuted dienophiles provided more intriguing results
(Scheme 6 and Table 4). Compounds 4 and 6 undergo
cycloaddition with both methyl methacrylate 14 and
methyl crotonate 15. However, although the reaction
with gem-substituted 14 affords a mixture of the 8-endo
and the 8-exo cycloadducts 16 and 17, reaction with vic-
substituted 15 affords a mixture of the 8-endo and the
7-endo cycloadducts 18 and 19. No traces of the other
regioisomeric cycloadducts were observed in the crude
We speculate that the differences between the reactivity
of the 6-methyl and the 6-phenyl substituted 2(H)-1,

7124
K. Afarinkia et al. / Tetrahedron Letters 45 (2004) 7121–7124
tern, indicating that the regiochemical preferences in
the cycloadditions are not exclusively determined by
the electronic demand of the dienes.
Acknowledgements
We thank Oxford GlycoSciences (UK) Ltd (now part of
Celltech R&D Ltd) for financial support (A.B.).
Supplementary data
Scheme 7. Cycloadditions of 5 and 7 with disubstituted dienophiles.
Supplementary data associated with this article can
be found, in the online version, at doi:10.1016/j.tetlet.
2004.07.105.
Table 5.
Dienophile
Diene
Ratio of cycloadductsa (yield %)b
References and notes
8-endo
7-endo
8-exo
7-exo
14
14
5
7
90 (32)
64 (50)c
0
0
10 (3)
36 (16)c
0
0
1. Afarinkia, K.; Bahar, A.; Neuss, J.; Ruggiero, A. Tetra-
hedron Lett. 2004, 45, 3995.
a Ratios of cycloadducts are normalised.
2. De Borggraeve, W.; Rombouts, F.; Van der Eycken, E.;
Hoornaert, G. J. Synlett 2000, 713.
b Yields after purification by silica gel chromatography.
c 33% of unreacted starting material was also isolated.
3. Vanaken, K. J.; Lux, G. M.; Deroover, G. G.; Meerpoel,
L.; Hoornaert, G. J. Tetrahedron 1996, 52, 2591–2602.
4. Vanaken, K. J.; Lux, G. M.; Deroover, G. G.; Meerpoel,
L.; Hoornaert, G. J. Tetrahedron 1994, 50, 5211–5224.
5. Tutonda, M. G.; Vandenberghe, S. M.; Van Aken, K. J.;
Hoornaert, G. J. J. Org. Chem. 1992, 57, 2935–2937.
6. Fannes, C.; Meerpool, L.; Hoornaert, G. J. Synthesis
1992, 705.
7. Meerpool, L.; Hoornaert, G. J. Synthesis 1990, 905.
8. Afarinkia, K.; Bahar, A.; Neuss, J. Synlett 2003, 2341.
9. Wu, X. J.; Toppet, S.; Compernolle, F.; Hoornaert, G. J.
Tetrahedron 2000, 56, 6279–6290.
4-oxazin-2-ones with di-substituted dienophiles 14 and
15 may be due to the influence of steric factors. Thus,
the lack of reactivity between 15 and 5 or 7 can be attrib-
uted to the repulsion between the sterically demanding
phenyl substituent at the 6-position of 2(H)-1,4-oxa-
zin-2-one azadiene with the b-methyl substituent in
methyl crotonate. Similarly, the preference for forma-
tion of the 7-endo cycloadduct in the cycloadditions
between 15 and 4, or 6, can be attributed to the greater
repulsion between the C-1 and C-7endo methyl groups,
which would have arisen in the 8-exo cycloadduct.
10. Wu, X. J.; Dubois, K.; Rogiers, J.; Toppet, S.; Compern-
olle, F.; Hoornaert, G. J. Tetrahedron 2000, 56,
3043–3051.
11. Afarinkia, K.; Nelson, T. D.; Vinader, M. V.; Posner, G.
H. Tetrahedron 1992, 48, 9111.
12. Woodward, B. T.; Posner, G. H. Adv. Cycloadd. 1999, 5,
47.
13. Afarinkia, K.; Daly, N. T.; Gomez-Farnos, S.; Joshi, S.
Tetrahedron Lett. 1997, 38, 2369–2372.
14. Afarinkia, K.; Posner, G. H. Tetrahedron Lett. 1992, 33,
7839–7843.
These results also show that although the pattern of
reactivity of 2(H)-1,4-oxazin-2-ones broadly follows
those of 2(H)-pyran-2-ones, there are subtle but signifi-
cant differences between them. In particular, although
2(H)-1,4-oxazin-2-ones can be considered as the 2-aza-
diene analogues of 2(H)-pyran-2-one dienes, they are
significantly more versatile and reactive species than
2(H)-pyran-2-ones.
15. Posner, G. H.; Nelson, T. D.; Kinter, C. M.; Afarinkia, K.
Tetrahedron Lett. 1991, 32, 5295–5298.
16. Afarinkia, K.; Berna-Canovas, J. Tetrahedron Lett. 2000,
41, 4955–4958.
17. Afarinkia, K.; Bearpark, M.; Ndibwami, A. J. Org. Chem.
2003, 68, 7158–7166.
18. Posner, G. H.; Wettlaufer, D. G. Tetrahedron Lett. 1986,
27, 667.
19. Posner, G. H.; Nelson, T. D.; Kinter, C. M.; Johnson, N.
J. Org. Chem. 1992, 57, 4083.
20. Posner, G. H.; Wettlaufer, D. G. J. Am. Chem. Soc. 1986,
108, 7373.
21. Posner, G. H.; Haces, A.; Harrison, W.; Kinter, C. M. J.
Org. Chem. 1987, 52, 4836.
22. Posner, G. H.; Kinter, C. M. J. Org. Chem. 1990, 55, 3967.
23. Posner, G. H.; Nelson, T. D. Tetrahedron 1990, 46, 4573.
In summary, we have demonstrated that 2(H)-1,4-oxa-
zin-2-ones are highly reactive azadienes that undergo
efficient regio- and stereoselective cycloadditions with
mono-substituted dienophiles. The presence of elec-
tron-donating groups at the 3-position on the 2(H)-
1,4-oxazin-2-one rings improves the regioselectivity of
the cycloadditions with electronically matched (i.e., elec-
tron deficient) dienophiles, however, it does not preclude
cycloadditions with electronically mis-matched (i.e.,
electron rich) dienophiles. Furthermore, cycloadditions
of electron rich 2(H)-1,4-oxazin-2-ones with disubstitut-
ed dienophiles 14 and 15, do not follow the same pat-