502
A. A. Provatas et al.
Table 3. Photolytic stability of FD&C and D&C dyes
RT, room temperature
DyesA, conditions
FD&C green #3 FD&C blue #1 D&C violet #2 Carmoisine
D&C green #8 D&C yellow #10
1 h, RT, 10 mM NaBH4, no BSA
No bleaching
No bleaching
No bleaching
Bleached
Low bleaching Low bleaching No bleaching No bleaching No bleaching
Bleached Bleached No bleaching No bleaching No bleaching
Low bleaching Low bleaching No bleaching No bleaching No bleaching
Bleached Bleached No bleaching No bleaching No bleaching
Bleached
No bleaching No bleaching No bleaching
1 h, RT, 10 mM NaBH4, 3.5 mM BSA
1 h, RT, 10 mM NaBH4, 10 mM Et3N, no BSA
1 h, RT, 10 mM NaBH4, 10 mM Et3N, 3.5 mM BSA No bleaching
1 h, hn, 10 mM NaBH4, 10 mM Et3N, no BSA No bleaching
1 h, hn, 10 mM NaBH4, 10 mM Et3N, 3.5 mM BSA No bleaching
Low bleaching Low bleaching No bleaching No bleaching No bleaching
AConcentration of dye: 1.0 mM.
hν/visible
illustrated that the co-presence of the photosensitizer, sacrificial
reductant, and sodium borohydride was essential in this process.
Although slight bleaching of the dye was observed, it is believed
∗
PS
PS
Ϫ
SET
ϩ
∗
ϩ
ϩ
PS ϩ Et3N
PS
Et3N
that the function of the photosensitizer dye is catalytic. This was
O
O
Ϫ
O
proven by a control experiment conducted in the absence of dye
that did not lead to the conversion of the epoxide under light
irradiation. UV–Visible spectra of the reaction mixture before
and after the photolysis showed only slight bleaching of the
photosensitizer. As mentioned above, BSA was added to assist
this catalytic process. Sodium borohydride was believed to
efficiently function as a radical scavenger of the radical inter-
mediate. As a result, sodium borohydride minimized the forma-
tion of side products. The regioselectivity that this method
provides depends on the radical anion formed from the ring
opening of the radical anion of the epoxide. It is believed that the
more stable radical is formed, leaving the hydroxyl group on the
less substituted carbon atom.
Ϫ
PS
R1
H
PS ϩ R1
H
R2
R3
R2
R3
H
Ϫ
O
Ϫ
R1
H
R1
R2
R3
R2
R3
ϩ
H
Ϫ
OH
H
O
H
H
R1
R1
H
R2
R3
R2
R3
Conclusions
Scheme 2.
In summary, a mild, practical, and convenient method of
regioselective photocleavage of epoxides has been developed
and can be used as an alternative to existing methods. The
alcohol product with the hydroxyl group on the less substituted
carbon atom was isolated in high yield and regioselectivity.
High conversions of the epoxide to the corresponding alcohol
(up to 100 %) were observed. Control experiments illustrated
that the presence of the photosensitizer dye, sodium boro-
hydride, and light was crucial. Furthermore, the presence of
triethylamine in the photoreaction mixture was necessary to
increase the reaction rate and percentage yield of the corre-
sponding alcohol. The presence of BSA and the use of non-toxic
D&C dyes as photosensitizers in the photoreaction mixture gave
promising results. Bleaching of the dyes was observed, but the
use of BSA reduced this effect. The combination of MG, car-
moisine, and D&C yellow #10 as photosensitizers along with
triethylamine, sodium borohydride, and BSA was needed to
achieve a high isolated yield of the corresponding alcohol in a
large scale set-up. In comparison, UV irradiation of the epoxide
substrate gave a high conversion to the alcohol but the photo-
reaction mixture did not appear to be clean as determined by gas
chromatography mass spectrometry (GC–MS) . In the presence
of visible light, light absorption only occurs by the catalytic dye.
In the UV-assisted photocleavage, if a complex substrate is
being irradiated, there is a chance of light absorption by one of
the possible chromophores present, leading to the initiation of
unwanted photochemical reactions.
the proper non-toxic dyes that would not bleach under the
photoreaction conditions, experiments were performed in a
9 : 1 methanol/water solution in the absence or presence of
BSA for each of the following dyes: FD&C (employed in food,
drugs, and cosmetics) green #3, FD&C blue #1, D&C violet #2,
carmoisine, D&C green #8, and D&C yellow #10, and methy-
lene green (MG).[16] The results are shown in Table 3.
Results from Table 3 indicated that dyes that could be used as
photosensitizers in this epoxide cleavage study were FD&C
green #3, carmoisine, D&C green #8, and D&C yellow #10.
Carmoisine, D&C green #8, and D&C yellow #10 were used in
Reactions 10, 11, 12, and 15 as shown in Table 2.
Based on the performed experiments, a mechanism for the
cleavage of epoxides under these conditions is envisaged, as
illustrated in Scheme 2. The proposed mechanism is similar to
that proposed by Hasegawa et al.[11] and Cossy et al.[12] to
explain the UV light-induced cleavage of epoxides. The initial
step is the excitation of the photosensitizer (PS) in the presence
of visible light. The dye-sensitized photocleavage of the epoxide
is achieved by interaction of the excited dye (PS*) with the
sacrificial reductant (triethylamine). The product of such single-
electron transfer (SET) interaction is the photosensitized radical
anion (PSꢀꢁ) that can react with the epoxide to give an epoxide
radical anion and the ground state PS. Cleavage of the epoxide
radical anion, and Hꢀ abstraction from the solvent may then
occur. Then, the Hþ abstraction from the solvent may lead to the
alcohol product. Control experiments (Reactions 8, 9, and 13)
The proposed epoxide photocleavage method was speci-
fically developed to use visible light from inexpensive
sources (sunlight). The purpose of this study was to introduce