Journal of the American Chemical Society
Article
evacuation, nearly no signal was observed for pristine CdS,
while new peaks at 1140 and 1108 cm−1 were observed for
increased the oxidation potential of the photoinduced holes
and serves as the proton acceptor, which accounts for the
enhanced photoactivity. The present work reports the pivotal
effect of surface sulfate ion on electron−proton transfer in
photocatalysis provides a novel and facile method for
increasing photocatalytic efficiency.
[
SO ]/CdS, which is ascribed to the adsorbed methoxy
4
41
groups. When 3-phenylpropionaldehyde was adsorbed on
the [SO ]/CdS and irradiated, the carbonyl stretching
4
−
1
frequency red-shifted from 1718 to 1712 cm , accompanied
group at 2723 cm− (Figure S13). These results demonstrate
that the sulfate ion serves as a proton acceptor and facilitates
the photoinduced dissociation of the O/C−H bond.
We then studied the adsorption of glycerol on the catalyst
1
EXPERIMENTAL SECTION
■
Preparation of [SO ]/CdS. CdS nanorods were synthesized by a
4
42
27
modified solvothermal method according to the literature. In a
typical procedure, 4.62 g of cadmium nitrate and 4.62 g of thiourea
were placed in a Teflon-lined steel chamber and filled with
ethylenediamine, followed by 160 °C treatment for 24 h. After
cooling to room temperature, the resulting solids were collected by
centrifugation and washed with deionized water. The products were
then dried at 60 °C overnight. CdS nanorods were finally obtained as
a yellow powder. The as-prepared CdS nanorods were then treated by
radio frequency excited plasma at room temperature for the desired
(Figure S3a). Glycerol floats above the pristine CdS(001) slab
without obvious interactions, while [SO ]/CdS(001) shows a
4
strong interaction with glycerol via dissociative adsorption with
an adsorption energy of −1.42 eV. The oxygen atom in [SO ]
4
abstracts H in the hydroxyl group and the remaining part
adsorbs on the S atom. DFT calculation shows that the
S3b). The adsorption of substrate on S atom favors the ET
process between the substrate and the holes.
time to afford [SO ]/CdS when the vacuum was kept around 0.2
4
mbar with a flow of air at 10 mL/min, and the power was kept as 200
W.
Characterizations. Crystalline information was measured by
powder X-ray diffraction (XRD) patterns, conducted on a PANalytical
X-Pert PRO diffractometer by using Cu Kα radiation at 40 kV and 20
mA. Continuous scans were in the 2θ range from 10°−80°. The
surface structure of the obtained catalysts was identified by X-ray
photoelectron spectroscopy (XPS) analyses, performed on a
ThermoFischer, ESCALAB Xi+. The binding energy was referenced
to the C 1s peak at 284.80 eV. The catalyst morphology
characteristics were observed by STEM images using a Titan G2
60−300. UV−vis DRS were recorded on a UV−vis spectropho-
tometer (UV-2600) at room temperature in the range of 300−900 nm
Lastly, a mechanism was proposed by combining the
reaction results and DFT calculation (Figure 5). Generally,
38,39
photoreaction involves a PCET process.
The photo-
induced holes serve as the electron acceptor, and the surface
sulfate ion [SO ] serves as the proton acceptor. The C−H/O−
4
H bond cleavage probably follows the PCET process. First,
upon photoirradiation, glycerol was dissociatively adsorbed on
the surface via the photoinduced PCET process as
demonstrated by the adsorptive FTIR characterization. The
electron in the O−H bond is transferred to the hole and the
with BaSO as the background. The surface functional group of the
4
obtained catalysts was identified by FTIR spectra performed on a
Thermo Scientific Nicolet iS10.
proton is abstracted by surface sulfate ion [SO ]. The proton is
4
reduced to a hydrogen atom by the photoinduced electron and
Adsorption FTIR. The methanol adsorption FTIR spectra were
recorded using a Thermo Scientific Nicolet iS10 IR spectrometer. The
spectra of the adsorbed methanol molecules have subtracted the
spectra of the samples before the adsorption. Methanol-adsorption
FTIR spectra were conducted as follows: the CdS samples were
placed in a homemade IR cell and evacuated (P < 10− Pa) at 423 K
for 0.5 h. Then, methanol vapor was introduced into the cell at 303 K
the surface sulfate ion [SO ] is restored. The couple of
4
hydrogen atoms will generate H . Glyceraldehyde was
2
generated via further C−H bond cleavage. Cleavage of O
C−H bond in glyceraldehyde will generate acyl radical via
PCET process like the O−H bond cleavage which was
S14). Photo irradiation of 3-phenylpropionaldehyde in the
presence of styrene produced addition coupling products,
which demonstrated that the acyl radical was formed during
3
−
2
and left for 0.5 h. The cell was then evacuated for 0.5 h (P < 10 Pa)
to remove the physically adsorbed methanol. Then methanol vapor
was again introduced into the cell at 303 K and left for 0.5 h upon
−2
photoirradiation with 450 nm laser (300 mW cm ). The cell was
photoirradiation. The surface sulfate ion [SO ] promotes the
4
−2
then evacuated for 0.5 h (P < 10 Pa) to remove the physically
Finally, decarbonylation of the acyl radical generates CO and
ethylene glycol (EG). The adsorbed EG further undergoes a
adsorbed methanol.
DFT Calculation Settings. All of the first-principles electronic
structure calculations were carried out using the Vienna ab initio
4
3
simulation package (VASP), one density functional theory
implementation. The exchange correlation potential was described
similar process to form CO and H . The reaction potential of
2
4
4
each step was calculated by the DFT. The reaction is
endothermic and is thermally unfavorable with a positive
reaction energy of 1.51 eV. The abstraction of the proton is an
exothermic reaction, whereas proton reduction to free
hydrogen is endothermic, which may account for the relatively
by the Perdew−Burke−Ernzerhof (PBE) formulation of the
generalized gradient approximation (GGA). The ion−electron
interactions were represented by the projector augmented wave
45
(
PAW) method. A plane wave basis set with an energy cutoff of 400
eV was used. The k-point sampling was performed using the
46
Monkhorst−Pack scheme. The electronic self-consistent minimiza-
lower H /CO ratio.
−5
2
tion was converged to 10 eV, and the geometry optimization was
converged to −0.02 eV. The lattice constants of CdS were optimized
CONCLUSIONS
In summary, we developed a method to generate surface ion
to be a = 4.200 Å, b = 4.191, and c = 6.817 Å, in good agreement with
■
[
47−49
the experimental constants, a = 4.14 Å and c = 6.72 Å.
We used
them to build a (2 × 2√3) CdS(001) slab with 10 atomic layers and
a vacuum of 15 Å. Atoms in the bottom 6 atomic layers were fixed to
their bulk positions, while the rest were allowed to fully relax. A 4 × 2
× 1 k-point mesh was used.
Reaction Procedure and Product Analysis. The reaction was
carried out in homemade LED photoreactors. Typically, 10 mg of
substrate and 5 mg of catalyst were added into 1 mL of solvent in a
SO ] on CdS via plasma treatment in air. Compared with
4
pristine CdS, [SO ]/CdS exhibits 9-fold higher CO generation
4
−
1
−1
rate (0.3 mmol g h ) and 4-fold higher H generation (0.05
2
−
1
−1
mmol g h ) in the photoreforming of glycerol. The [SO ]/
4
CdS is stable and could be used for nearly 150 h without loss
of activity. Detailed studies demonstrate that surface ion [SO4]
6
539
J. Am. Chem. Soc. 2021, 143, 6533−6541