I. Funes-Ardoiz et al. / Tetrahedron 69 (2013) 9766e9771
9767
level of the photoisomerization, the comparison with the well-
known GFP chromophore and the design of modifications for
new switches.
4) position. However, when having an electron-withdrawing group,
such as nitro (switch 2e, entry 5), there is no significant change in
the isomers ratio compared with 2b. Alternatively, if there is
a bromine (switch 2a, entry 1), the percentage of E isomer at the
PSS also increases. Moreover, if substituent R1 is modified from Me
(switch 2a, entry 1) to Ph (switch 2h, entry 8) and R2 remains the
same, the percentage of E isomer at the PSS significantly increases.
Therefore, the value of the isomers ratio at the PSS can vary
depending on the substituents R1 and R2, which represent a great
advantage. Even more important is that all of the values for the
isomers ratio have improved with respect to the values obtained for
each switch when irradiated with a 125-W medium-pressure Hg
lamp since the percentage of E isomer has increased in all cases. In
addition to this, there was no decomposition of the switches when
irradiated under these conditions, so no side reactions occurred
during the irradiation time intervals, which is a further betterment
upon the previous conditions used.
Irradiating under these improved conditions, mixtures of Z and
E isomers of different compounds bearing electron-withdrawing
groups were subsequently separated to individually characterize
and study them. With this aim, we irradiated a 0.01 M solution of
compound 2a (100% Z isomer at t¼0 min) in acetonitrile in a pho-
toreactor with lamps (14 lampsꢀ8-W/lamp) with emission wave-
length centred at 350 nm, until the PSS was reached. Then, the
resulting mixture of isomers was separated by flash chromatogra-
phy on silica gel, using hexane/ethyl acetate (10:1) as eluent. We
then studied the kinetics of the isomerization reaction under these
conditions. The photoisomerization reaction was followed by 1H
NMR at short irradiation intervals. Showing the isomers ratio ver-
sus irradiation time, we obtained the two graphs in Fig. 1(a) cor-
responding to the sample with 100% of Z-2a isomer at the
beginning of the reaction, and (b) corresponding to the sample with
100% of E-2a isomer at tirrad¼0 min.
2. Results and discussion
2.1. Wavelength dependence
Benzylideneeoxazolones 2 are good candidates for efficient
photoswitches as they are easily synthesized, feature good photo-
isomerization quantum yields and are thermally stable.9 We have
previously shown that the irradiation wavelength affects the pho-
tostationary state (PSS) in this type of compounds. Also, in our
preliminary tests, we detected the presence of side-products in the
reaction crude in some cases when using a medium-pressure Hg
lamp as the standard light source. These competing reactions took
place when an electron-withdrawing group was placed in the
phenyl of R2. Thus, we aimed for the use of light of different
wavelengths in order to diminish the contribution of side reactions.
Therefore, we irradiated at 360 nm a 2.2ꢀ10ꢁ3 M solution of
compound 2a (R1¼Me, R2¼p-BrPh) in CDCl3 in a quartz cuvette
with a monochromator, and followed the photoisomerization
process by 1H NMR, until the photostationary state was reached.
The value for the isomers ratio at the PSS was 54% (Z)/46% (E),
which greatly differed from the result obtained when irradiating in
a 125-W medium-pressure Hg lamp and a Pyrex filter [85% (Z)/15%
(E)] used in our previous reaction conditions.9 The different ab-
sorption of the E and Z isomers in these diverse irradiation condi-
tions explains the experimental outcome. In addition, there were
no signals in the 1H NMR spectra of side reactions so the photo-
isomerization process was taking place exclusively. This implies
a considerable improvement over the previous irradiation condi-
tions used in the case of compounds yielding partial de-
composition. Therefore, we decided to repeat the irradiation
process of different compounds using a light source with emission
wavelength centred at 350 nm to compare the two irradiation
methods. For this purpose, we prepared 0.01 M solutions of dif-
ferent photoswitches 2 in acetonitrile, placed them in Pyrex test
tubes in a carrousel in a photoreactor, and irradiated them using
lamps with emission wavelength centred at 350 nm (14 lampsꢀ8-
W/lamp) until the PSS was reached. Depending on the compound,
the irradiated mixture took from 1 to 2 h to reach the PSS and the
reaction was followed by 1H NMR. The results are shown in Table 1.
As shown in Fig. 1, the value of the isomers ratio at the PSS for
these two processes is 51% (Z)/49% (E), which is reached after ca.
85 min of irradiation. In addition, we can calculate the values for
the kinetic constants of each process from the first points of each
graph, and therefore determine the value of the relative kinetic
constant that relates the rate of both processes. If we consider
kZ/E¼1, then kE/Zz1, which means that the E/Z isomerization
process is almost as fast as the Z/E reaction, so there is no dif-
ference in rate.
In addition, there are slight differences in the absorption spectra
with solvents of different polarities. Thus, to determine if there is
any noticeable effect in the values of the isomers ratio at the PSS,
we irradiated 2a in different solvents. Discernible differences in the
values for the isomers ratio at the PSS were found (see Table S1).
However, the differences between these results were unrelated to
the polarity of the solvents according to the varying polarity indices
Table 1
Photostationary state for molecular photoswitches 2 irradiated at 350 nm
Entry
R1
R2
Compound
Ratio at PSS
% Z
% E
reported in literature, such as ENT ,
3 , or
p
(Taft).19
*
1
2
3
4
5
6
7
8
Me
Me
Me
Me
Me
Me
Me
Ph
p-BrPh
p-Tol
2a
2b
2c
2d
2e
2f
2g
2h
2i
54
75
42
41
76
62
70
36
40
80
55
46
25
58
59
24
38
30
64
60
20
45
p-MeOPh
o-MeOPh
p-NO2Ph
p-CNPh
Ph
p-BrPh
o-Br
p-NO2Ph
p-CN
2.2. Photophysics
We have previously reported9 that these photoswitches are
quite efficient as the quantum yield of the reaction using the two
isomers of 2g were found to be VZ/E¼0.25ꢂ0.01 starting from the
Z isomer and VE/Z¼0.11ꢂ0.02 from the E isomer. Different de-
activation processes, such as fluorescence among others could di-
minish the efficiency of these photoswitches. Therefore, we
measured the emission and excitation spectra of two illustrative
photoswitches: the two isomers (Z and E) of compound 2g (R1¼Me,
R2¼Ph, see Fig. 2), and compound 2a (R1¼Me, R2¼p-BrPh) (only the
Z isomer). A 3.6ꢀ10ꢁ5 M solution of Z-2g in deoxygenated aceto-
nitrile at 298 K was measured in a spectrofluorimeter. Only a weak
emission band centred at 380 nm was obtained when exciting from
250 to 340 nm (Fig. 2).
9
10
11
Me
Ph
Ph
2j
2k
The following trends can be inferred from the results in Table 1.
If the substituent of the phenyl group in R2 is an electron-donor
group, such as methoxy (switch 2c, entry 3), the percentage of E
isomer at the PSS increases with respect to 2b. The effect is the
same whether the substituent is in para or ortho (switch 2d, entry