Inorganic Chemistry
Article
(dmgH)2]2+. This rate is also significantly slower than the
1.23 × 109 M−1 s−1 quenching rate between photoexcited
RePS* and [Co(dmgH)2]2+, thus excluding this mechanism.64
The absence of new transient features and long-lived signals
in the TRIR spectra of 2f with [Co(dmgH)2]2+ (independent
of the presence of TEOA) may suggest that either (i) a fast
back electron transfer from the cobalt complex in the absence
of TEOA or (ii) a fast electron transfer from TEOA (present in
large excess) to 2f+ takes place after the initial quenching event
in the three-component system.
Combining the insights gained from these experiments, we
believe that the formation of an exciplex between excited 2f*
and TEOA can simultaneously explain the changes in the
emission spectra and the unchanged TRIR data. The formation
of a metal complex−molecule exciplex has been observed
before,66 with the case of [Re(4,7-diphenyl-1,10-
phenanthroline)(CO)3Cl] and N,N-dimethylaniline (DMA)
in decalin being particularly relevant in the context of our
work. In this system, even very low concentrations (∼10−3 M)
of DMA were enough to observe a new red-shifted emission
band.67 While our specific scenario is differentnamely, a
more polar solvent and a different quenchercomplex 2f has
very large charge-transfer character upon photoexcitation (as
shown in Figure 5), leading to a large change in the dipole
moment, which would render the formation of a charge-
transfer exciplex with TEOA feasible.
Figure 11. Hydrogen evolution from 2f as a photosensitizer.
Conditions: [2f] = 50 μM, [Co] = 0.5 mM, [dmgH2] = 3.5 mM, 1
M TEOA, 0.1 M TfOH in DMF, and λexc = 453 nm.
fine-tuning of the conditions of the photocatalytic experiments,
which are outside the scope of the present work.
To gain a deeper understanding of the initial photochemical
steps in the photocatalytic cycle and, in particular, to evaluate
the potential differences between the ILCT and MLCT
excited-state reactivities, we performed emission quenching, as
well as TRIR experiments in the picosecond-to-microsecond
time scales of complexes 2d and 2f in the presence of TEOA.
For 2d, we observed a very small change in the excited-state
kinetics and emission intensities, leading to a diffusion-limited
quenching rate kQ ≈ 4 × 108 M−1 s−1. We find this number of
qualitative character, given the very short lifetime of 2d, which
limits the possibility of studying the intermolecular electron-
transfer steps in this complex.
Additional mechanistic studies of these complexes with the
purpose of unravelling the nature of this process and its solvent
dependence will reveal further differences in the reactivity and
3
3
photophysical properties of the ILCT versus MLCT excited
states, easily switched in these complexes by remote
substitution in the ligand framework.
Complex 2f, in contrast, presented an unexpected behavior.
Upon the addition of TEOA, we observed the appearance of a
strongly emissive blue-shifted band whose intensity increased
with increasing concentrations of TEOA. The absorption
spectra remained unchanged. At the same time, the spectral
features and kinetics of 2f (1 mM in degassed DMF) remained
largely unchanged in a concentration range from 1 mM to 1 M
TEOA (containing 10 mol % TfOH), hinting at a different
order of events in the photocatalytic cycle of 2f. TRIR spectra
recorded in the absence of TfOH also did not show any
difference with respect to those of pure 2f in DMF.
CONCLUDING REMARKS
■
The ground and excited states of a series of rhenium(I)
tricarbonyl complexes with substituted 4′-(4-R1-phenyl)-
2,2′:6′,2″-terpyridine ligands were analyzed by steady-state
and time-resolved spectroscopic methods. We found excellent
correlations between the evaluated spectroscopic and electro-
chemical properties of these complexes and the Hammett σp
substituent constants, showing their tunability even by
changing the substituent in a remote position of the ligand
framework. CV and IR-SEC revealed an irreversible stepwise
two-electron reduction, which, in turn, allowed us to assign the
identities and spectral signatures of the one- and two-electron-
reduced species.
3
In the “conventional” photocycle, the MLCT excited state
of a rhenium(I) complex (like RePS) is reductively quenched
in a bimolecular fashion by TEOA, and it is this reduced
species that transfers the electrons to the proton reduction
catalyst.14,64 If the 3ILCT excited state of 2f is not quenched by
TEOAand given the fact that the system does evolve
significant amounts of hydrogenwe hypothesize that this
complex transfers the electron directly to the cobalt catalyst
(considering its very long lifetime). After this electron transfer,
the oxidized 2f+ is regenerated by TEOA at longer time scales.
To prove this hypothesis, we performed TRIR experiments
of 2f with different concentrations of [Co(dmgH)2]2+ ranging
from 10 to 75 mM (the cobalt complex was prepared as a stock
solution from anhydrous CoBr2 and 3 equiv of dmgH2 in
DMF). From Stern−Volmer analysis of the TRIR kinetic data
(Figure S11), we obtained a second-order quenching rate
The NMe2-substituted complex (2f) showed notably
exceptional lifetime and spectroscopic properties compared
to other complexes of the same series. In this complex, the
lowest singlet and triplet excited states both have ILCT
character, as demonstrated by TRIR and emission spectra, and
are supported by detailed (TD-)DFT calculations. The
particularly long triplet lifetime of ca. 380 ns obtained for
the NMe2-substituted complex (vs ca. 1.5 ns for the
unsubstituted complex) suggests that very strongly donating
groups can lead to distinct photophysical and photochemical
properties and opens new avenues to exploring further
substituent-related modifications to the terpyridine ligand
framework. We conclude that judicious ligand-based sub-
stitutions can still be exploited to access long-lived, strongly
emissive yet sufficiently reducing excited states of complexes
with significantly red-shifted absorption, which would perform
as better photosensitizers by using a wider portion of the solar
constant of kCo = (2.0
0.2) × 108 M−1 s−1 for electron
transfer between 2f and [Co(dmgH)2]2+. This value of kCo
compares very well with the reported values of 1.3 × 108 and
2.5 × 108 M−1 s−1 for electron transfer between reduced
RePS− or [Re(bpy)(CO)3Br]−, respectively, and [Co-
G
Inorg. Chem. XXXX, XXX, XXX−XXX