M.-J. Wei, et al.
JournalofPhotochemistry&PhotobiologyA:Chemistry387(2020)112137
there was no signal of the low spin (S = 1/2) CoII species occurred to
both two samples, disclosing the weak 6-coordination in Co-CP. While
exposing Co-CP to visible light for photocatalysis, an intense new signal
with a g value of 3.49 (at around 120 m T) was observed, which is
characteristic of an octahedral high-spin (S = 3/2) CoII species. The
photo-induced EPR signal indicated that there was charge transfer in
the Co-CP during photocatalysis [29].
The cyclic voltammetry (CV) measurement was performed on a
DMF solution of 0.1 M Bu4NPF6 and the Co-CP coated electrode, which
revealed the photo-induced electron transition inside the catalyst. As
shown in Fig. 3d, there was two irreversible reduction peaks at -2.21 V
and -1.47 V, respectively, which could be ascribed to the signal electron
transfer from [CoIICoII] to [CoICoII] and then [CoICoII] to [CoICoI] [16].
It was generally believed that in the photocatalytic process, the pho-
tosensitizer [Ru(bpy)3]2+ underwent an excited state [Ru(bpy)3]2+* to
quenching process, possibly by oxidative quenching or reduction
quenching pathway [18]. In particular, the excited state [Ru(bpy)3]2+
*
Fig. 3. a) PL emission spectra of Co-CP and H4L ligand. b) XPS spectra of Co-CP.
c) CW EPR spectra of pristine Co-CP and the Co-CP after photocatalytic hy-
drogen production. d) The Cyclic voltammogram of Co-CP. Measured condi-
tions: Ag/AgCl was used as the reference electrode, indium-tin oxide (ITO) glass
electrode was used as the working electrode, and a platinum plate was used as
counter electrode; electrolyte, a DMF solution of Bu4NPF6 (0.1 M); scan rate,
0.1 V/s.
might be oxidized by the Co-CP catalyst, forming [Ru(bpy)3]3+ species,
or might be reduced by the sacrificial agent TEOA to form [Ru(bpy)3]1+
species, both of which accompanied with an electron-transfer (ET)
process. In attempt to verify the quenching pathway of [Ru(bpy)3]2+*,
the photoluminescence measurements of original Ru(bpy)3Cl2 solution
and the solution with added 1 mL of TEOA were carried out (Fig. S11).
There was no fluorescence quenching but a fluorescence enhancement
occurred while adding 1 mL of TEOA into original mixed solution, in-
dicating an oxidatively quenching pathway for the excited state [Ru
(bpy)3]2+*. Additionally, comparing the UV–vis diffuse reflectance
spectra of Co-CP and the emission spectrum of [Ru(bpy)3]2+, partial
spectra overlap illustrated a feasible energy-transfer (EnT) from [Ru
(bpy)3]2+ to the catalyst Co-CP (Fig. S12) [18].
DMF/H2O (4/1, v/v) mixed solvent (Fig. S7). Through a series of
control experiments, it was found that triethanolamine (TEOA) worked
best as a sacrificial agent (Fig. S8). Screening the above conditions,
4 mg Co-CP, 4 mg Ru(bpy)3Cl2, and 1 mL TEOA were added into 5 mL of
DMF/H2O (4/1, v/v) pre-mixed solvent and exposed to visible light for
photocatalytic hydrogen evolution, which could reach an activity of
1778 μmol g-1 h-1 at 3 h. The Co-CP-catalyzed HER reaction was further
investigated by recycling experiments, in which Co-CP was centrifuged
from the catalytic reaction solution, washed with DMF for 3 times, and
redistributed in the fresh solvent after every 2 h illumination. As shown
in Fig. S9, Co-CP maintained catalytic activity within three cycles, in-
dicating that the catalyst is recyclable and stable for photocatalytic
hydrogen.
In attempt to study the separation and recombination of photo-in-
duced electron-hole pairs in the catalyst, photoluminescence (PL)
measurements were performed on Co-CP and H4L ligand. Both excited
by 460 nm, H4L ligand presented a maximum emission peak at 575 nm,
while the PL of the Co-CP was almost quenched, showing the rapid
transfer of photo-generated electrons and lower recombination rate of
the photo-induced electron-hole (Fig. 3a). Furthermore, the photo-
current response of Co-CP was recorded versus Ag/AgCl under visible
light. As soon as the light was turn on, a fast and apparent current signal
was observed, while the signal decayed rapidly with the light turned
off. The fast, uniform and invertible photocurrent response shown in
Fig. S10 suggested the efficient separation of photo-generated charge
carriers in Co-CP.
Taking into account the above factors, we propose the possible
mechanism of Co-CP photocatalytic hydrogen production as following.
In the presence of the visible light, the photosensitizer [Ru(bpy)3]2+
was excited to from photoactive [Ru(bpy)3]2+*, which then underwent
an oxidatively quenching pathway accompanied by the generation of
[Ru(bpy)3]3+ species. The excited photoelectron formed from the
LUMO state of Ru(bpy)3Cl2 transfer to the unsaturated CoII sites. The
Co2+ in the catalyst Co-CP was reduced from the [Ru(bpy)3]2+* to
form an intermediate state [CoIICoI] or [CoIICoI], and the [Ru(bpy)3]3+
immediately obtained electron from sacrificial agent TEOA, achieving a
loop. Finally, the protons gathered on the catalyst surface by hydrogen
bonding react with the photoelectrons on the unsaturated CoII cites to
continuously generate hydrogen. In general, Co-CP provides a good
platform to promote the transfer of photoelectrons and inhibit the
combination of photo-generated electron-holes of the Ru(bpy)3Cl2. The
photocatalytic process is shown in Scheme 1.
High-resolution X-ray photoelectron spectroscopy (XPS) measure-
ment was used to probe the surface chemical state of cobalt in Co-CP. As
shown in Fig. 3b, the peaks at 781.3 and 797.5 eV were attributed to Co
2p3/2 and Co 2p1/2 core level photo emission. And the two adjacent
satellite peaks at 785.8 and 801.9 eV were originated from Co2+ shake-
up excitation [25,26], which indicating the existence of Co2+ species in
Co-CP. To elucidate the mechanism of the photocatalytic process,
continuous wave (CW) EPR spectroscopy was used to determine the
active intermediates formed during visible-light irradiation over the Co-
CP at 100 K. Recorded at X-band microwave frequency (9.85 GHz), the
CW EPR spectra of pristine Co-CP in the dark and after visible-light
driven photocatalytic hydrogen production were presented in Fig. 3c.
According to literature, the downfield signal at g = 2.04 both occurred
to the pristine Co-CP and the catalyzed sample can be attributed to the
spatially confined amino-groups in the Co-CP [27,28]. Furthermore,
Scheme 1. Plausible mechanism for visible-light driven photocatalytic hy-
drogen production by Co-CP.
4