Angewandte Chemie International Edition
10.1002/anie.201902766
COMMUNICATION
These observations indicated that the low quantum yields of 5
were related to molecular rotations. In contrast, BODIPY 1
exhibited high quantum yields (Table S4) and their emission
intensities were not sensitive to solvent viscosities (Figure 1h).
These results again corroborate our theoretical calculations.
We continued to perform transient absorption spectroscopy
measurements on 5 (Figures 1i and S38—S41). After photo-
excitation, our experimental data revealed the co-existence of
two excited states. The first excited state produced a simulated
emission band at ~480 nm, with a short decay lifetime of 3.0 ps
Finally, significant rotations of the dialkylated-amino moieties
and the associated fluorescence quenching in the TICS state
can be employed in many practical applications.[ Inspired by
the pioneering work of de Silva et al.,[5b] we are interested in
applying TICS compounds as fluorescent probes. Since the
shuttle mechanism involves rotations of molecules, they are
good candidates for quantifying solvent viscosities (such as by 5
and 15; Figures 1g and 2e). Since TICS compounds possess
low-polarity, they could be insensitive to solvent polarity, making
them dedicated viscosity probes.
14]
(
Figure 1j). We assigned this band to the LE state. The second
We also expected that replacement/removal of the amino
group at the meso-position could eliminate TICS and enhance
emission intensities. Indeed, the reaction between 15 and
glutathione (GSH) removed the meso-aniline group and turned
on bright red emissions peaked at 611 nm (Figure 2f). In
addition, our in vitro experiments showed that 15 was highly
selective to GSH (Figures S34—S37). Inspired by these
successes, we next proceeded to perform live cell experiments.
By staining HeLa cells with 5 μM Probe 15 and 3 μM Hoechst
33342 (a nucleus stain) for 2 h, we observed bright emissions
from both the red channel and the blue channel (due to the
presence of GSH and nucleus, respectively; Figure 2g). In
contrast, by pre-treating cells with 1 mM N-Methylmaleimide
(NMM) to remove GSH, we observed no red emission in this
control group, but only blue emissions from the nucleus (Figure
2h). These experiments demonstrated that TICS compound 15
is a promising GSH probe for both in vitro and in vivo
experiments. Interestingly, we also noted many existing NO and
phosgene fluorescent probes actually belonged to the TICS
family (Figures S23 and S24).
In conclusion, we have discovered a new charge transfer and
separation process, namely twisted intramolecular charge
shuttle (TICS), through chemical database searches, quantum
chemical rationalizations, and experimental validations. In TICS
compounds, the dialkylated-amino groups acted as an electron-
withdrawing group upon light absorption but became an
electron-donating group after the ~90° intramolecular rotation.
This unique role switching and “charge shuttle” process
differentiated TICS from the TICT mechanism. We also
demonstrated that TICS was applicable to a wide range of
chemical families of fluorophores and could be employed to
construct useful fluorescent probes. The discovery of the TICS
mechanism provides an expansive view on charge transfer and
separation processes existing in nature and will inspire potential
applications.
excited state showed an excited state absorption band with a
peak and a shoulder at 555 nm and 521 nm, respectively, and a
much longer decay lifetime of 78.3 ps (Figure 1k). Note that in
the TICS state, a negatively charged BODIPY radical will form
upon the complete charge transfer from the amino group to the
BODIPY scaffold. It was reported that this BODIPY radical has a
transient absorption spectrum at ~550—580 nm (Figure S42).[13]
This is in good agreement with our observation. We thus
attributed the second excited state to the TICS state. The
transition rate from the LE to the TICS states was estimated to
be 3.3×1011 s-1.
Encouraged by the consistency between theoretical and
experimental results, we next explored if the TICS model was
also applicable to oxygen-rhodamines (Figures 2a and S12—
S14). Our calculations showed that dialkylated amino group
substituted rhodamine 6 exhibited two stable conformation in the
1
S state, and the ~90° twisted conformation was the global
minimum (Figure 2b). Moreover, 6 clearly exhibited the “charge
shuttle” character: the amino group at the meso-position served
as an EWG in the FC and LE states but switched to an EDG in
the TICS state upon photoexcitation. Our calculations also
showed that the primary amino substituted rhodamine 7 is not
prone to TICS formation, as the twisted state is unstable in the
PES of 7 (Figure 2c).
We next synthesized 6 and 7 and measured their spectral
properties (Tables S5 and S6; Figures S30—S33). As
predicated by our theoretical calculations, the quantum yields of
6
was very low due to substantial TICS formations, while 7
emitted bright fluorescence. These results also suggested that
the TICS mechanism was not limited to BODIPY dyes but could
exist in other dye families.
As such, we computationally explored if TICS was applicable
to various families of fluorophores. Our theoretical calculations
showed that TICS was energetically favourable in coumarin 8,
alkoxy BODIPY 9, oxygen-rhodamine 10 and 11, carbon-
rhodamine 12, and silicon-rhodamine 13 (Figures 2d and S15—
S20). We also found that in TICS compounds, the amino group
is not limited to dialkylated amino groups, but can be extended
to aniline groups as well, such as in BODIPY 14 and rhodamine
Acknowledgements
WC, RL and XL were indebted to the financial support from
SUTD and the SUTD-MIT International Design Centre (IDC)
1
5 (Figures S21 and S22). Overall, our results demonstrated
that TICS represented a general charge transfer and separation
process, governing many different types of fluorophores.
Among these predicted TICS compounds, we chose to
synthesize 15. Indeed, as solvent viscosity increased, the
emission intensities of 15 were greatly enhanced (Figure 2e).
These results confirmed the rotational nature of the TICS
compound 15.
[
T1SRCI17126, IDD21700101, IDG31800104]. QQ, WL and ZX
were supported by NSFC (21878286, 21502189) and DICP
DMTO201603, TMSR201601). YST, DG and MJL were
(
supported by NRF Singapore through SMART Centre’s BioSyM
IRG research program. YTC was supported by Institute for Basic
Science (IBS) [IBS-R007-A1]. The authors would like to
This article is protected by copyright. All rights reserved.