Angewandte
Chemie
Table S6), which indicates that the formation of {PtIICl2-
a monoradical species. Based on our published methods,[8]
(bpy) (A)nꢀ1(AꢀC)} is not thermodynamically favored for
deconvolution was carried out to determine the relative
abundances of the MV+C and (MV+C)2 species at each
irradiation time (Table S9). Consequently, the time course
of the number of electrons stored within a framework can be
viewed for all of these triads, as depicted in Figure 2B. For
both [PtCl2(4,4’-MV2)]4+ and [PtCl2(4,4’-MV4)]8+, the total
number of electrons shows a maximum at around 10 min and
gradually decreases as the reaction proceeds, which suggests
that an induction period is spent and then a stationary state is
established, during which the formation and consumption of
the radical species are balanced in rate, as previously
observed.[8] These results are also consistent with the induc-
tion periods observed in the H2 evolution profiles of these
triads in Figure 1A (see also the Supporting Information).
The behavior of [PtCl2(5,5’-MV4)]8+ is essentially similar to
those of the 4,4’-substituted derivatives, but the maximum
appears much earlier at around 3 min, which must be viewed
as related to the largest driving force of reductive quenching
being estimated for this system (see above). The apparent
dimerization constants (KD) are in the range of 103–105 mꢀ1
(Tables S8–S10) and are much larger than the KD values
reported for free MV2+ in aqueous media (KD = 550mꢀ1),[10]
which reveals that the present dipeptide backbone provides
an extraordinary stabilization effect for the diradical species.
The dependence of the initial H2 evolution rate on the
EDTA concentration obeys Michaelis–Menten kinetics for all
of these triads (Figures S19–S21). Importantly, no H2 evolves
in the absence of EDTA. This behavior is quite similar to
those previously observed for most of the PHEMDs devel-
oped so far in our group.[3–5] Saturation behaviors are also
seen when the PHEMD concentration is varied at a constant
EDTA concentration (Figures S22–24). In the same manner,
nanosecond transient absorption (TA) spectroscopy
(Figure 3) shows that a photoproduct attributable to {PtIICl2-
ꢀ
these systems. The E1/2(bpy/bpyꢀC) value is most positive
shifted for [PtCl2(5,5’-MP2)]2+ (ꢀ1.18 V vs. Fc/Fc+; Table S6).
For either MV2/MV4 or MP1/MP2 derivatives, the 5,5’-
substituted derivative shows the highest activity as a PHEMD
among each series. Therefore, it seems quite likely that the
driving force for the reductive quenching of the 3MLCT state
by EDTA governs the net photocatalytic efficiency.
Next, we pay attention to the number of electrons stored
over the multiviologen tethers within each framework. It is
expected that more than one reducing equivalent can be
transferred to the multiviologen tethers as a result of
successive reductive quenching reactions. As a typical exam-
ple, the absorption spectral changes during the H2 evolution
driven by [PtCl2(4,4’-MV4)]8+ are shown in Figure 2A. This is
(bpy) (A)nꢀ1(Aꢀ·)} is only observable in the presence of
ꢀ
EDTA. Moreover, for all of the triads, the yield of photo-
product monitored by TA spectroscopy exhibits saturation
with regard to the variation in the EDTA concentration
(Figure 3B; Figures S25 and S26). These results strengthen
Figure 2. A) Spectral changes during the photolysis of an aqueous
acetate buffer solution (pH 5.0; at 208C under Ar) containing 0.1 mm
[PtCl2(4,4’-MV4)]8+, 0.1m NaCl, and 30 mm EDTA. B) Changes in the
total number of electrons stored within a framework during photolysis.
2ꢀ
our conclusion that the electron injection from the YH2
species ion paired with the cationic PHEMD is the key to
II
afford the initial photoproduct {Pt Cl2(bpy) (A)nꢀ1(AꢀC)}.
ꢀ
Finally, in situ dynamic light scattering (DLS) measure-
ments were carried out to ascertain the lack of colloidal
platinum dispersion during the H2 evolution photocatalyzed
by these multivilogen-tethered PHEMDs. As shown in
Figure 4, variations in the light-scattering intensity for all of
these triads during the photolysis were negligible and the
development of any exponential decay feature in the corre-
lation functions was also negligible (Figures S27–S29). These
results clearly show that the formation of colloidal platinum
particles can be ruled out during the H2 evolution photo-
catalyzed by these triads. For comparison, Figure 4 shows the
manner in which the light-scattering intensity increases when
the same amount of a well-known colloid precursor (K2PtCl4)
is converted into colloidal platinum particles under the same
experimental setup. These results clearly indicate that the
photocatalytic H2 generation driven by these multiviologen-
similar to the charge-storage behavior previously observed
for photoinduced H2 evolution driven by a combination of
[Ru(bpy)3]2+ and [PtCl2(MVn)] (n = 2, 4, or 6; MVn denotes
5,5’-substituted-bpy ligands tethered to 2, 4, and 6 viologen
units) in the presence of EDTA,[8] for which oxidative
quenching of [Ru(bpy)3]2+* by [PtCl2(MVn)] was assumed to
be the major electron-transfer event. In other words, we were
unware of the activity of [PtCl2(MVn)] as a PHEMD in that
study. As previously described,[8] the 900 nm band corre-
sponds to a diradical species, which occurs as a consequence
of specific stabilization of a stacked dimer (MV+C)2 within the
same residue. The spectral features of this diradical species
are quite consistent with those previously reported by Lee
et al.[10] and also with those previously observed for [PtCl2-
(MVn)].[8] On the other hand, the 602 nm band corresponds to
Angew. Chem. Int. Ed. 2014, 53, 4618 –4622
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4621
Kitamoto, Kyoji
