M. Blanco-Lomas et al. / Tetrahedron Letters 55 (2014) 3361–3364
3363
Table 2
stabilization for the cationic species. Table 3 shows some of the
data obtained (see Supplementary material).
PSS and relative ratios for neutral and quaternized switches
**
Compound
E isomer* (%)
Z isomer* (%)
kmet/neutral
A qualitative agreement between the experimental and the
computed bands was found. In all cases, the computed bands are
red-shifted by 20–30 nm with respect to the experimental bands.
Again, the protonated and methylated derivatives show the same
behaviour, with the low-energy absorption band caused by the
HOMO–LUMO transition (see Supplementary material). In all
cases, this transition implies the displacement of electron density
from the HOMO (mainly located at the R1 substituent) to the LUMO
(located at the pyrroline moiety). In addition, in order to under-
stand the different thermal behaviour of 1 and 4, we also computed
the transition structures (TS’s) for the thermal conversion (Fig. 2,
see Supplementary material).
1
5
2
6
3
7
4
8
70
90
56
68
34
40
48
75
30
10
44
32
66
60
52
25
1.5
1.4
1.1
1.2
*
Isomers ratio at the photostationary state. For the protonated compounds 9–11
the same results were obtained.
**
Relative kinetic constants (kmet/neutral).
Interestingly, both TS’s share similar geometrical features
although small differences can account for the diverse experimen-
tal outcome. In both cases, the TS features a twisting along the cen-
tral C@C bond (74° vs 57°) that causes the molecule to divide into
two halves. The second aromatic ring in 4 causes an extended con-
jugation in the naphthyl moiety that leads to an increased C–C dis-
tance. This fact, in turn, causes a bigger destabilization in the
structure as the initial C@C bond is more distorted. Thus, a TS
higher in energy was found (57.6 kcal/mol for 1, 71.8 kcal/mol for
4) causing that the thermal back-reaction is hampered when R1
is naphthyl. It should be noted that the computed energy barriers
are too high. This may be due to the performance of the theoretical
level. The energy barriers for the methylated derivatives 5 and 8
follow the same trend with barriers of 53.4 kcal/mol and
69.6 kcal/mol, respectively.
they are quaternized. From the linear relationship between the
first points of each graph for each pair of compounds (representing
ln[A0]/[A] vs irradiation time and doing a linear fit), we can calcu-
late the values for the kinetic constants of each process, and there-
fore, the value of the relative kinetic constant that relates the rate
of both processes (see Table 2). In all cases, being the kinetic con-
stant for the neutral compounds kneutral = 1, we noticed that the
E?Z isomerization process for the quaternized compounds is
approx. 1.5 times faster than the E?Z reaction for the neutral
compounds. Therefore, methylation of the photoswitches caused
only a mild acceleration of the isomerization reaction rate.
We then checked the thermal stability of the quaternized com-
pounds. We measured the back-reaction of the irradiated samples
for both the methylated and protonated derivatives. For the deriv-
atives of 5–7, the reversion to the thermodynamically stable E iso-
mer takes place within hours. Clearly, these results would imply a
serious drawback in practical applications that require a delayed
energy liberation at rt. Interestingly, for 8 and 11 no back reaction
was detected by 1H NMR in 6 days. However, the E isomer recovery
could be easily performed by heating the sample at 60 °C. In this
case, the energy barrier prevents relaxation from the higher to
the lower energy isomer, exactly as the functioning of the MOST
systems require.
Conclusions
We have shown that small modifications in the general struc-
ture of a series of rhodopsin-based molecular photoswitches can
alter their properties to match the requirements of MOST systems.
Interestingly, the nitrogen atom quaternization and the substitu-
tion can be used to independently modify both the regions on
the spectrum in which these molecules absorb and the barrier
for reversion. This approach has proven useful in the preparation
of compounds subject to taking part in technological devices. Fur-
ther studies to design, synthesize and characterize new candidates
for MOST systems are currently underway.
In order to understand the changes made in the chromophore
upon quaternization we performed some density functional theory
calculations on the tested compounds. This will be helpful also in
the rational design of new compounds with specific properties.
We started by computing 1 using B3LYP with the standard basis
set 6-31G* and acetonitrile as the solvent through the Polarizable
Continuum Model (PCM) method (see Supplementary material).
The inclusion of solvent is not critical for the neutral compounds,
but it is required to take into account the effect of charge
Experimental section
The methylated photoswitches were synthesized as follows: To
a solution of the neutral photoswitch (1 equiv) in dry toluene
(2 ml), under nitrogen atmosphere, was added methyl triflate CF3-
SO3CH3 (1 equiv). The resulting mixture was stirred vigorously for
Table 3
Experimental and computed absorption bands (nm)
Compound
Experimental
Computed
1
5
9
2
6
3
7
10
4
289
334
334
304
371
335
337
337
311
324
350
415
350
415
305
360
359
326
393
388
370
366
318
340
370
438
370
439
8
11
Figure 2. Computed transition structures and displacement vectors for the thermal
conversion for 1 (left) and 4.