The effect of solvent polarity on the product distribution in the reaction of singlet oxygen with enolic tautomers of 1-(2',4',6'- trialkylphenyl)-2-methyl 1,3-diketones

Full text of DOI:10.1021/jo991024v

J . Org. Chem. 1999, 64, 9247-9250
9247
Notes
compounds 2 (Scheme 1).6 These reactions proceed by the
abstraction of hydrogen from the enolic hydroxy group.
We report here the effect of solvent polarity on the singlet
oxygenation of enolized 1-aryl-2-methyl 1,3-diketones
that the abstraction of hydrogen from the enolic hydroxy
group decreases substantially with the increase of ab-
straction from the 2-methyl group as the solvent polarity
decreases.
Th e Effect of Solven t P ola r ity on th e
P r od u ct Distr ibu tion in th e Rea ction of
Sin glet Oxygen w ith En olic Ta u tom er s of
1-(2′,4′,6′-Tr ia lk ylp h en yl)-2-m eth yl
1,3-Dik eton es
Michikazu Yoshioka,* Yuumi Sakuma, and
Masaichi Saito
Resu lts a n d Discu ssion
Department of Chemistry, Faculty of Science, Saitama
University, Shimo-okubo, Urawa, Saitama 338-8570, J apan
We have recently reported that the reaction of singlet
oxygen with enolized 1,3-diketones 3a -d using Methyl-
ene Blue (MB) as the sensitizer at room temperature in
acetonitrile gave the enedione 4a -d , the epoxy ketone
5a -d , and the hydroperoxy ketone 6a -d . In the photo-
oxygenation of 3b-d in methanol under the same condi-
tions, the initially formed enedione 4b-d reacted with
methanol to give the corresponding Michael adducts 7
(Scheme 2).7 In that paper, we proposed that the enedione
4 and the epoxy ketone 5 would arise from the initially
formed hydroxyperoxide 8 (Scheme 3). The hydroxyper-
oxide 8 decomposes to afford 4 with loss of hydrogen
peroxide. The hydrogen peroxide thus generated would
react with 4 to yield 5. In the present study, we tried to
detect the proposed intermediate 8 by low-temperature
experiments. After a carbon tetrachloride solution of 3c
had been photooxygenated below -20 °C until the start-
ing material was completely consumed, the solvent was
removed under vacuo below -20 °C. The NMR analysis
of the mixture revealed that the reaction proceeded
cleanly to give a single product that could be stored below
-20 °C but transformed quantitatively into 4c on stand-
Received J une 28, 1999
The reaction of singlet oxygen with olefins has been
extensively studied and a number of comprehensive
reviews have been published.1 The olefins with allylic
hydrogen atoms undergo ene-type reactions to form
allylic hydrogenperoxides. In the ene reaction with
substituted olefins, hydrogen abstraction by singlet oxy-
gen takes place preferentially at the more highly substi-
tuted side of the double bond (cis effect).2 For cis olefins
with a bulky alkyl group, a preference for hydrogen
abstraction on the larger alkyl group has been observed.3
The weakly electrophilic singlet oxygen readily reacts
with highly substituted olefins. By contrast, the reactions
with electron-deficient systems such as R,â-unsaturated
carbonyl compounds are often slow but show a marked
regioselectivity.4 In these systems, hydrogen abstraction
takes place selectively on an alkyl group geminal to the
electron-withdrawing group. However, the reaction of
singlet oxygen with enolic tautomers of 1,3-dicarbonyl
compounds, â-hydroxy-R,â-unsaturated carbonyl com-
pounds, has received little attention. The reaction of
singlet oxygen with enolic tautomers of 1,3-diketones in
the presence of fluoride ion, which enhances the nucleo-
philicity of enols, gives ketonic products resulting from
R-hydroperoxide formation.5 The enolic tautomers 1 of
acyl- and alkoxycarbonylbenzocycloalkenones react with
singlet oxygen to give 2-hydroperoxy 1,3-dicarbonyl
1
ing at room temperature for several hours. The H NMR
spectrum of the initially formed product at -20 °C
showed two singlets at δ 6.22 and 6.59 due to olefinic
protons. These peaks gradually disappeared with the
gradual appearance of two singlets assignable to olefinic
protons of 4c at δ 5.89 and 6.36 on standing at room
temperature. The 13C NMR spectrum of the initially
formed product showed a carbonyl carbon peak at δ 207.3.
1
However, the H NMR spectrum could not fully explain
(1) For general reviews of the reaction of singlet oxygen, see: (a)
Frimer, A. A. Chem. Rev. 1979, 79, 359. (b) Singlet Oxygen; Wasser-
man, H. H., Murray, R. W., Eds.; Academic Press: New York, 1979.
(c) Stephenson, L. M.; Grdina, M. S.; Orfanopoulos, M. Acc. Chem. Res.
1980, 13, 419. (d) Singlet Oxygen; Frimer, A. A., Ed.; CRC press:
Florida, 1985. (e) The Bimolecular Reactivity of Singlet Molecular
Oxygen in Advances in Photochemistry; Volman, D. H., Hammond, G.
S., Neckers, D. C., Eds.; Wiley: New York, 1992; Vol. 17, pp 217-274.
(2) (a) Orfanopoulos, M.; Grdina, M. B.; Stephenson, L. M. J . Am.
Chem. Soc. 1979, 101, 275. (b) Schulte-Elte, K. H.; Rautenstrauch, V.
J . Am. Chem. Soc. 1980, 102, 1738.
the structure of 8c for the initially formed product. The
methine proton of isopropyl group appeared at δ 2.1-
2.2, fairly high field compared to that of the isopropyl
1
ketone. The H NMR spectra of 4c, 5c, and 6c showed
the methine proton of the isopropyl ketone group at δ
3.26, 3.02, and 4.38, respectively.7 On the basis of the
NMR data together with chemical behavior, the structure
of the initially formed product may be assigned as 9c.
Although the hydroxyperoxide 9c was transformed
quantitatively into the enedione 4c, the epoxy ketone 5c
could not be detected on standing the CDCl3 solution in
an NMR tube at room temperature for a long time.
(3) Orfanopoulos, M.; Stratakis, M.; Elemes, Y. Tetrahedron Lett.
1989, 30, 4875.
(4) (a) Ensley, H. E.; Carr, R. V. C.; Martin, R. S.; Pierce, T. E. J .
Am. Chem. Soc. 1980, 102, 2836. (b) Orfanopoulos, M.; Foote, C. S.
Tetrahedron Lett. 1985, 26, 5991. (c) Adam, W.; Griesbeck, A. Angew.
Chem., Int. Ed. Engl. 1985, 24, 1070. (d) Adam, W.; Griesbeck, A.
Synthesis 1986, 1050. (e) Kwon, B.-M.; Kanner, R. C.; Foote, C. S.
Tetrahedron Lett. 1989, 30, 903. (f) Adam, W.; Richter, M. J . Tetra-
hedron Lett. 1993, 34, 8423.
(6) Yoshioka, M.; Nishioka, T.; Hasegawa, T. J . Org. Chem. 1993,
58, 8, 278.
(5) Wasserman, H. H.; Pickett, J . E. J . Am. Chem. Soc. 1982, 104,
4695.
(7) Yoshioka, M.; Hashimoto, K.; Fukuhara, T.; Hasegawa, T. J .
Chem. Soc., Perkin Trans. 1 1998, 283.
10.1021/jo991024v CCC: $18.00 © 1999 American Chemical Society
Published on Web 11/11/1999

9248 J . Org. Chem., Vol. 64, No. 25, 1999
Notes
Ta ble 1. Regioselectivity in th e Rea ction of Sin glet
Oxygen w ith 3a
Sch em e 1
distributionb
regioselectivityc
sub-
strate solvent sensitizer
4
5
6
7
(4 + 5 + 7)
6
3a
3a
3a
3a
3b
3b
3b
3b
3b
3c
3c
3c
3c
3c
3d
3d
3d
3d
3d
CCl4
TPP
TPP
TPP
MB
TPP
TPP
TPP
MB
31 69
17 79
100
96
88
70
100
91
70
59
65
100
85
79
58
67
100
80
77
60
58
benzene
acetone
MeCN
CCl4
benzene
acetone
MeCN
MeOH
CCl4
benzene
acetone
MeCN
MeOH
CCl4
benzene
acetone
MeCN
MeOH
4
4
12
30
3
84 12
44 26 30
75 25
26 65
14 56 30
45 14 41
9
9
30
41
35
MB
6
25 40
Sch em e 2
TPP
TPP
TPP
MB
75 25
42 43 15
47 32 21
15
21
42
33
49
9
7
42
27 47
MB
TPP
TPP
TPP
MB
100
80
77
20
23
40
37 52
20
23
40
42
60
MB
a
A solution of 3(150 mg) in a solvent (100 mL) in the presence
of sensitizer (ca. 10 mg) was irradiated with a 100 W tungsten-
halogen lamp through a K2CrO4 filter solution until the starting
b
material was totally consumed. Determined by 1H NMR on the
fractions after chromatography of the mixture, integrating olefinic
hydrogens of 4, epoxymethylene hydrogens of 5, and a hydroperoxy
hydrogen of 6. In the photooxygenation in metanol, the isolated
yields were given. c Compounds 4, 5, and 7 are products arising
by hydrogen shift from the methyl group and the compound 6 is
that arising by hydrogen shift from the hydroxy group.
of 3 in polar solvents gave the hydroperoxy ketone 6 in
substantial amounts together with 4 and 5. The propor-
tion of 6 increased with decreasing proportion of the sum
yield of the enedione 4 and the epoxy ketone 5 as the
solvent polarity increased. These products were fully
characterized as described previously.7 The 1H NMR
spectrum of 4 showed two singlets at δ 5.8-5.9 and 6.4-
6.6 due to two olefinic protons and that of 5 showed
methylene protons of the epoxy ring at δ 2.9-3.2. The
hydroperoxy group of 6 appeared at δ 8.9-9.9. The
product ratios of the photooxygenation of 3, which were
determined by 1H NMR spectroscopy on the fractions
after chromatography of the mixture, are given in Table
1. The ratios of 4/5 were dependent on the experimental
conditions. The proportion of 4 increased with decreasing
proportion of 5 when the mixture was passed through a
silica gel column after standing at room temperature for
several hours because the initially formed hydroxyper-
oxide 9 was gradually transformed into 4 on standing at
room temperature but converted into 5 on treatment with
silica gel.
Sch em e 3
However, when the hydroxyperoxide 9c was treated with
silica gel immediately after irradiation at low tempera-
ture, it was quantitatively transformed into the epoxy
ketone 5c. In this case, no enedione 4c was detected. On
the contrary, the same treatment of the CDCl3 solution
after the hydroxyperoxide 9c had been completely con-
verted into the enedione 4c with loss of hydrogen
peroxide with silica gel resulted in partial transformation
of 4c to 5c. These results suggest that the epoxy ketone
5c was produced from the hydroxyperoxide 9c by both
intramolecular epoxidation and intermolecular process
between the enedione 4c and generated hydrogen per-
oxide, though the role of silica gel was not at all clear.
To examine the solvent effect of the reaction, we
studied the photooxygenation of 3a -d in three other
solvents, namely carbon tetrachloride, benzene, and
acetone. The product ratios were dependent on solvent
polarity. Photooxygenation of 3 in carbon tetrachloride
using tetraphenylporphine (TPP) as the sensitizer at
room temperature gave 4 and 5. The hydroperoxy ketone
6 could not be detected. However, the photooxygenation
Perepoxide is a possible intermediate of the singlet
oxygenation reaction.8 In our study, four perepoxides A,
B, C, and D are possible because the starting 3 exists in
solution in two cis-enolic forms owing to intramolecular
hydrogen bonding7,9 (Scheme 4). The two enolic forms are
in rapid dynamic equilibrium. The perepoxide C in which
the terminal oxygen atom orients toward the methyl side
would be expected to be transformed into the hydroxy-
(8) (a) Inagaki, S.; Fukui, K. J . Am. Chem. Soc. 1975, 97, 7480. (b)
Dewar, M. J . S.; Thiel, W. J . Am. Chem. Soc. 1977, 99, 2338. (c) Grdina,
B.; Orfanopoulos, M.; Stephenson, L. M. J . Am. Chem. Soc. 1979, 101,
3111. (d) Hotokkna, M.; Roos, B.; Siegbahn, P. J . Am. Chem. Soc. 1983,
105, 5263. (e) Orfanopoulos, M.; Smonou, I.; Foote, C. S. J . Am. Chem.
Soc. 1990, 112, 3607.
(9) (a) Gorodetsky, M.; Luz, Z.; Mazur, Y. J . Am. Chem. Soc. 1967,
89, 1183. (b) Iglesias, E. J . Phys. Chem. 1996, 100, 12592.

Notes
J . Org. Chem., Vol. 64, No. 25, 1999 9249
Sch em e 4
examining the transition states of this reaction. In the
transition state TsA, the precursor to the perepoxide A,
the oxygen is placed anti to the hydroxy and carbonyl
groups, and the net dipole moment is expected to be
smaller than that in the transition state TsB, the
precursor to the perepoxide B, where the oxygen has a
syn orientation with respect to the hydroxy and carbonyl
groups (Scheme 5). Therefore, the more polar transition
state TsB is expected to be stabilized better by polar
solvents than the less polar transition state TsA.12 The
solvent dependence of side selectivity of the ene reaction
of singlet oxygen with R,â-unsaturated esters is explained
in terms of the stability of perepoxide-like transition
states.12 In the transition state TsA, singlet oxygen
interacts with both allylic hydrogen atoms on C2 and C3.
The effective interaction between the hydroxyl group and
the partially negatively charged singlet oxygen in the
transition state TsB requires the coordination of the
hydroxyl group with the incoming singlet oxygen.13 In
nonpolar solvents, the intramolecular hydrogen bonding
between the hydroxy and carbonyl groups prevents such
coordination. In polar solvents, the intramolecular hy-
drogen bonding must compete with intermolecular hy-
drogen bonding with the solvent. Consequently, the
hydroxyl group can interact with the incoming oxygen
by hydrogen bonding through a favorable five-membered
ring.14 Thus, TsB is stabilized in polar solvents than in
nonpolar solvents, reflecting the increased proportion of
the hydroperoxy ketone 6.
Sch em e 5
peroxide 8 by a hydrogen shift from the methyl group to
the negative oxygen. Similarly, the perepoxide A would
be transformed into the hydroxyperoxide 9. However,
since the structure of the initially formed hydroxyper-
oxide was assigned as 9 and not 8, the perepoxide C could
be excluded from the possible intermediate. Therefore,
the perepoxide D may not intervene. The photochemical
reaction of benzoylacetone with olefins forms [2 + 2]
adducts that are derived largely from the enolic form in
which the oxygen adjacent to the phenyl ring is in the
enol form.10 On the contrary, the reaction of singlet
oxygen with 3 exclusively took place on the enolic form
in which the oxygen adjacent to the aromatic ring is in
the keto form. Since the C1-C2 site of 3 is sterically more
congested than the C2-C3 site owing to the bulky 2,4,6-
trialkylphenyl group on C1, the favored attack of singlet
oxygen would take place on the C2-C3 site. Although the
enedione 4 and the epoxy ketone 5 are produced from
the perepoxide A through the hydroxyperoxide 9, the
perepoxide A may be also converted into the hydroperoxy
ketone 6 by the hydrogen shift from the alkyl group on
the carbon bearing the hydroxyl group or by open of the
ring to a cation E followed by proton shift (Scheme 5).
However, in the photooxygenation in carbon tetrachlo-
ride, no hydroperoxy ketones 6 were detected. The
geminal selectivity observed in R,â-unsaturated carbonyl
system also operates in our system. The perepoxide B
would be transformed into the hydroperoxy ketone 6
probably via the cation E. The enol ethers undergo ene-
type reaction with singlet oxygen via the perepoxide
intermediate, which opens the ring to the cation stabi-
lized by the presence of ether functionality.11
Exp er im en ta l Section
1H NMR spectra were recorded at 200, 300, and 400 MHz
using tetramethylsilane as an internal standard with CDCl3 as
solvent. 13C NMR spectra were recorded at 100 MHz with CDCl3
as solvent. Starting materials were prepared by previously
reported methods.15
Gen er a l P r oced u r e for th e P h otooxygen a tion of 3. A
solution of the particular substrate (150-500 mg) and ca. 10
mg of tetraphenylporphine or Methylene Blue in 100 mL of a
variety of solvents was irradiated with a 100 W tungsten-
halogen lamp through an aqueous solution of K2CrO4 (0.27 g
L-1) and Na2CO3 (1 g L-1) at room temperature while a stream
of air was passed continuously through the reaction mixture.
The course of the reaction was monitored by means of TLC; after
all starting material had been consumed, the solvent was
evaporated and the residue was fractionated by silica gel column
eluting with a mixture of hexane and ethyl acetate (6:1 to 10:
1). The composition of each fraction was determined by 1H NMR
(see Table 1 for product ratios). Further purification by column
chromatography afforded pure samples of enediones 4, epoxy
ketones 5, and hydroperoxy ketones 6. The physical properties
of these products have been already reported.7
Low -Tem p er a tu r e P h otooxygen a tion of 3c. A solution of
3c (34 mg) and a catalytic amount of tetraphenylporphine in
100 mL of CCl4 was irradiated below -10 °C by the same
manner as described above. After 3c had been consumed, the
solvent was removed under reduced pressure at -20 °C to give
the hydroxyperoxide 9c: 1H NMR (400 MHz, -20 °C) δ 0.99
(3H, d, J ) 7 Hz), 1.06 (3H, d, J ) 7 Hz), 2.1-2.2 (1H, m), 2.19
(6H, s), 2.32 (3H, s), 6.22 (1H, s), 6.59 (1H, s), 6.88 (2H, s), 8.07
(1H, s), and 10.31 (1H, s); 13C NMR (100 MHz, -20 °C) δ 15.5
The observed solvent effect on the selectivity of the
reaction of singlet oxygen with 3 can be rationalized by
(12) Orfanopoulos, M.; Stratakis, M. Tetrahedron Lett. 1991, 32,
7321.
(13) (a) Adam, W.; Nestler, B. J . Am. Chem. Soc. 1992, 114, 6549.
(b) Adam, W.; Nestler, B. J . Am. Chem. Soc. 1993, 115, 5041. (c) Prein,
M.; Adam, W. Angew. Chem., Int. Ed. Engl. 1996, 35, 477.
(14) Adam, W.; Bru¨nker, H.-G.; Kumar, A. S.; Peters, E.-M.; Peters,
K.; Schneider, U.; Schnering, H. G. J . Am. Chem. Soc. 1996, 118, 1899.
(15) Yoshioka, M.; Funayama, K.; Hasegawa, T. J . Chem. Soc.,
Perkin Trans. 1 1989, 1411.
(10) Casals, P.-F.; Ferard, J .; Ropert, R. Tetrahedron Lett. 1976,
3077.
(11) (a) J efford, C. W.; Rimbault, C. G. J . Am. Chem. Soc. 1978,
100, 6437. (b) Kwon, B.-M.; Foote, C. S. J . Org. Chem. 1989, 54, 3878.
(c) Chan, Y.-Y.; Li, X.; Zhu, C.; Liu, X.; Zhang, Y.; Leung, H.-K. J . Org.
Chem. 1990, 55, 5497.

9250 J . Org. Chem., Vol. 64, No. 25, 1999
Notes
(q), 17.4 (q), 19.3 (q), 21.2 (q), 36.9 (d), 107.4 (s), 128.0 (d), 128.2
(d), 133.2 (s), 134.3 (s), 135.4 (s), 138.0 (t), 138.8 (s), 144.2 (s),
and 207.3 (s).
of 9c into the enedione 4c. This solution was poured onto silica
gel in hexane, and the mixture was stirred at room temperature
for 2 h. The solvent was evaporated after removal of silica gel.
The 1H NMR spectrum showed that the residue was composed
of a 2:3 mixture of the enedione 4c and the epoxy ketone 5c.
Ch em ica l P r op er ties of th e Hyd r oxyp er oxid e 9c. The
hydroxyperoxide 9c was poured onto silica gel in a 5:1 mixture
of hexane and ethyl acetate, and the mixture was stirred at room
temperature for 4 h. The solvent was evaporated after removal
of silica gel by filtration. The 1H NMR analysis revealed the
quantitative transformation of 9c into the epoxy ketone 5c. The
hydroxyperoxide 9c was dissolved in CDCl3 in an NMR tube,
and the tube was left standing at room-temperature overnight.
Su p p or t in g In for m a t ion Ava ila b le: 1H NMR and 13C
NMR spectra of 9c. This material is available free of charge
via the Internet at http://pubs.acs.org.
1
The H NMR analysis revealed the quantitative transformation
J O991024V