Full Paper
tial polarization. In agreement with that model are the small
substituent effects on the absorption energy for HTI 5 versus
HTI 2 observed theoretically and the experimentally found
weak substitution effects on the absorption energy for HTIs
Approaching the limit of HTI photoisomerization rate
The limit for the donor–acceptor approach to increase the
photoisomerization rate can now be estimated quantitatively
+
1
to 5. Even widely different donor strengths (Br to OMe, that
+
by the Hammett s value at which the linear regressions of
is, s =0.15–À0.78) do not change the absorption energy as
shown in Figure 4. When the excited HTI 2 quickly moves out
of the FCS1Z region, the initial polarization persists until a local
minimum (MinS1Z) is reached (Figure 6a). Again, substituents at
the stilbene fragment should not influence the MinS1Z, because
they do not affect the polarization. The MinS1Z can be probed
directly by measuring stationary fluorescence. Experimentally
observed substituent effects on the stationary fluorescence
energy are indeed small for HTIs 1–5 (see Figure 4).
the two subsets of data (HTIs 1–5 and HTIs 6–8; Figure 3a,b)
+
meet. A s value of À1.1 was found consistently for both the
Z/E and E/Z direction. By using an appropriate donor substitu-
+
ent with that s constant it should be possible to achieve the
maximum rate of photoisomerization for HTIs with a lifetime
of only 1.2 ps for the Z/E direction and a lifetime of 0.6 ps for
the E/Z direction.
Instead of using a single substituent to approach the limit
for HTI photoisomerization rate, it is also possible to use more
than one medium strong donor group at the stilbene fragment
of the molecule. Thus, HTI 9 was prepared, bearing one ortho-
and one para-OMe group. This substitution leads to a Z/E pho-
toisomerization time constant of 2.4 ps, which is as fast as for
Before the molecule can leave the MinS1Z region and enter
the conical intersection (CoIn) to reach the ground state (S0),
a barrier has to be overcome (red bars indicate the barrier
heights in Figure 6). This barrier is the result of a crossing be-
tween two excited-potential surfaces: the S state and another
NH -substituted HTI 6. Similarly, an increase of the E/Z photo-
1
2
excited state (S ) with charge-transfer character. The S excited
isomerization time constant to 1.4 ps is observed for HTI 9, al-
ready exceeding the E/Z photoisomerization rate of HTI 6.
Hence, HTI 9 shows the fastest photoisomerization of a HTI re-
ported to date. To approximate the extent of donating ability
2
2
state is strongly polarized with a positive charge on the stil-
bene fragment. This polarized structure of the S excited state
2
can be stabilized by donor substituents at the para-position of
+
the stilbene fragment leading to a lowering of the energy. In
of the two OMe substituents in HTI 9, the Hammett s con-
+
contrast, the S state with the positive charge at the sulfur
stant of the para-OMe substituent was doubled, yielding a s
1
atom is not affected by these donor substituents. The substitu-
value of À1.56. This donor strength should already lead to a de-
celeration of the photoisomerization kinetics and a significant
redshift of the absorption and fluorescence; this is not ob-
served in this case. Apparently, multiple weak donor substitu-
ents at the stilbene fragment of HTIs can be used to approach
and possibly even exceed the rate limit of photoisomerization
set by a single strong donor. However, for the analogous
double substitution with SMe substituents (HTI 10), no in-
crease of the photoisomerization rate was detected.
ent induced stabilization of the S state versus the unaffected
2
S1 state reduces the barrier of photoisomerization as shown
schematically in Figure 6b for HTI 5. Thus, the photoisomeriza-
tion is accelerated with increasing donor strength of the sub-
stituent at the stilbene fragment as found experimentally for
HTIs 1–5 (Figure 3).
When the very strong donor groups NH , NMe , and juloli-
2
2
dine are used (HTIs 6–8), a different behavior occurs. Now a sig-
nificant stabilization of the S state is observed, leading to di-
1
minished absorption energies (Figure 4). The stabilization of
the FCS1Z structure in HTIs 6–8 is explained by the change of
Conclusion
the electronic character of the S state. The para-amine sub-
1
stituents stabilize the positive partial charge generated during
photoexcitation in the FCS1Z region. Similarly, a significant stabi-
lization of the MinS1Z structure is observed for HTIs 6–8, as can
be seen from the Hammett plots of the stationary fluorescence
energy (Figure 4). Again, localization of the partial positive
charge at the amine substituent is responsible for the ob-
Rational design of HTI photoswitches to enable the broad ap-
plicability in, for example, biology or supramolecular chemistry
is still limited by incomplete understanding of the photophysi-
cal properties. To gain detailed insight into substitution effects
on the photoswitching behavior, we synthesized a series of
substituted HTIs and analyzed the photoisomerization rates,
absorption and fluorescence profiles, quantum yields, as well
as thermal bistability. By increasing the donor strength of the
substituents at the stilbene fragment, we were able to signifi-
cantly increase the photoisomerization rate of the HTI photo-
switches. However, we found that there is a limit to the
donor–acceptor approach, beyond which an increase in the
substituent’s donating capacity does not further increase the
photoisomerization rate. Instead, HTIs with very strong donor
substituents, such as NH , NMe , or julolidine, at the stilbene
served stabilization of the MinS1Z. The effect of S excited-state
2
stabilization on lowering the barrier for photoisomerization is
thus counterbalanced by S state stabilization. This leads to an
1
increase of the barrier instead of a further decrease (Figure 6c).
The unexpected deceleration of the photoisomerization rate
with stronger donor substitution of HTIs 7 and 8 can therefore
be explained in terms of concomitant stabilization of the S1
and the S excited states. Thus, a rate limit for photoisomeriza-
2
tion of donor-substituted HTI photoswitches is established.
A similar mechanism is operative for the corresponding E/Z
photoisomerization, leading again to a rate limit for this photo-
isomerization direction with strong donor substituents at the
stilbene fragment.
2
2
fragment follow an opposite trend where stronger donor ca-
pacity leads to slower photoisomerization. Taking both trends
into account, we were able to identify a principal upper limit
of the photoisomerization rate for donor-substituted HTIs. This
Chem. Eur. J. 2014, 20, 13984 – 13992
13989
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim