Visual observation of hydrogen bubble generation from monodisperse CoP QDs on ultrafine g-C3N4 fiber under visible light irradiation

Article abstract of DOI:10.1039/c9ta09897f

Solar-driven hydrogen evolution reaction (HER) via water splitting is an attractive technology to address the growing demand for clean fuels. g-C₃N₄ is a promising candidate among photocatalysts, but it is plagued by its puny HER activity and miserly quantum efficiency. Tremendous efforts have been made to enhance g-C₃N₄ performance on HER; however, it is still far below the expectations in industrial production. Herein, we report a monodisperse CoP QDs-modified ultrafine g-C₃N₄ fiber (CoP/CNF) via in situ electrostatic adsorption deposition followed by low-temperature phosphatization treatment. The CoP/CNF showed an HER activity of 2.42 mmol h^-1 under visible light, 34.9 times higher than that of Pt/CNF, in which hydrogen bubbles evolution was observed with an apparent quantum efficiency of 59.9% at 420 nm. This benchmark HER activity was mainly because the CoP QDs could significantly suppress photoinduced charge recombination and improve the heterointerface HER rate. This work provides a useful strategy for designing highly active catalysts for solar-to-hydrogen fuel conversion.

Full text of DOI:10.1039/c9ta09897f

Journal of
Materials Chemistry A
View Article Online
View Journal | View Issue
PAPER
Visual observation of hydrogen bubble generation
from monodisperse CoP QDs on ultrafine g-C3N4
fiber under visible light irradiation†
Cite this: J. Mater. Chem. A, 2019, 7,
25908
*
*
Yingchun Xia, Wenjia Song and Shenglian Luo
Yunxiong Zeng,
Solar-driven hydrogen evolution reaction (HER) via water splitting is an attractive technology to address the
growing demand for clean fuels. g-C3N4 is a promising candidate among photocatalysts, but it is plagued by
its puny HER activity and miserly quantum efficiency. Tremendous efforts have been made to enhance g-
C3N4 performance on HER; however, it is still far below the expectations in industrial production. Herein, we
report a monodisperse CoP QDs-modified ultrafine g-C3N4 fiber (CoP/CNF) via in situ electrostatic
adsorption deposition followed by low-temperature phosphatization treatment. The CoP/CNF showed
an HER activity of 2.42 mmol hꢀ1 under visible light, 34.9 times higher than that of Pt/CNF, in which
hydrogen bubbles evolution was observed with an apparent quantum efficiency of 59.9% at 420 nm.
This benchmark HER activity was mainly because the CoP QDs could significantly suppress
photoinduced charge recombination and improve the heterointerface HER rate. This work provides
a useful strategy for designing highly active catalysts for solar-to-hydrogen fuel conversion.
Received 8th September 2019
Accepted 22nd October 2019
DOI: 10.1039/c9ta09897f
rsc.li/materials-a
meanwhile, the turnover number (TON) and turnover frequency
(TOF) reached ca. 3 270 000 in 90 h and 36 400 for Ni2P QDs,
Introduction
respectively. Inspired by such unprecedented progress, TMPs
are considered to be a promising alternative to precious metals
and noble-metal alloys for photocatalytic HER in the future.17
Photocatalysts are also equally pivotal to HER catalysis.18,19
Metal-free g-C3N4 is a distinctive photocatalyst due to its visible
light absorption and extremely low cost.16,17,20 Compared to the
typical transition-metal-based photocatalysts (TMBPs),16 g-C3N4
displays environmental pollution-free activity without being
oxidized or suffering from toxic ions leakage.21 To exploit a cost-
effective solar water-splitting system, integrating TMPs with g-
C3N4 into a hierarchical heterojunction could still be consid-
ered an ideal option.17,19 Hitherto, limited studies have focused
on TMPs/g-C3N4 composites, yet in those the HER performance
of TMPs/g-C3N4 was found to be fairly inferior to that of TMPs/
TMBPs.8 This situation is mainly caused by the following three
factors. First, g-C3N4 generally possesses a broader band gap
(ꢁ2.7 eV) than that of TMBPs, leading to lower light utiliza-
tion.17,19 Second, the photoinduced charge-recombination rate
of g-C3N4 is faster than that of TMBPs owing to the larger
carrier-migration resistance.22 Third, the absence of active sites
of TMPs leads to slow HER kinetics in TMPs/g-C3N4.23 Thus, the
fabrication of monodisperse TMPs with abundant active sites
on the nanostructured g-C3N4 with robust photocatalysis HER
activity and long-term stability is of great signicance.
Water splitting HER is of great interest because it can provide
fuel energy that is free from environment pollution.1,2 Oen,
high-performance catalysts, especially Pt-based catalysts, are
usually required for splitting water HER at a low energy barrier.3
The scalable consumption of Pt though is greatly hindered by its
high cost and scarcity.4 In this case, diverse non-precious
compound catalysts, such as transition-metal carbides
(TMCs),5 phosphides (TMPs),6 and chalcogenides (TMDs),7 have
been widely investigated as alternatives to Pt. However, these
transition metal composites still cannot achieve HER activity
matching that of Pt-group catalysts, except for TMPs.8
Usually, TMPs are a signicant class of composites derived
from the alloying of metals and phosphorus, and they have been
employed in electrochemical water splitting. In recent years,
a growing amount of TMPs have been exploited for HER, such
as Ni2P,9 FeP,6,10 Cu3P,11 MoP,12 WP2 13 and CoP,14,15 which have
shown competitive HER activity compared to Pt-based catalysts.
Although breakthroughs have been made to improve the HER
performance of TMPs,10,12,14 there are very limited investigations
on TMPs applied in photocatalytic HER compared to electro-
chemical HER. Ni2P QDs/CdS nanorods were reported with an
around 1200 mmol hꢀ1 gꢀ1 photocatalysis HER rate,16
State Key Laboratory of Chemo/Biosensing and Chemometrics, College of Chemistry &
Chemical Engineering, Hunan University, Changsha 410082, P. R. China. E-mail:
yxzeng@hnu.edu.cn; sllou@hnu.edu.cn
Herein, a CoP QDs/CNF heterogeneous nanojunction was
prepared via an in situ electrostatic adsorption deposition, fol-
lowed by low-temperature phosphatization treatment (Scheme
1). At rst, bulk g-C3N4 was oxidized into g-C3N4 quantum dots
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c9ta09897f
25908 | J. Mater. Chem. A, 2019, 7, 25908–25914
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
Journal of Materials Chemistry A
Co3O4/g-C3N4 nanober (Co3O4/CNF) and bare Co3O4
preparation
First, 50 mg g-C3N4 nanober was dispersed in a mixed solution
containing 50 mL absolute ethanol and 0.6 mL concentrated
aqueous ammonia. A certain amount of Co(OAc)2 solution (60.0
mM) was gradually dropped into the mixed solution and stirred
for 20 h at 80 ꢂC on a magnetic stirring apparatus simulta-
neously. Then the mixed solution was transferred to an 80 mL
Scheme
adsorption
1
Preparation of CoP/CNF via an in situ electrostatic
deposition followed by low-temperature
phosphorization.
a
ꢂ
Teon-lined autoclave and maintained at 150 C for 3 h. Aer
by the hydrothermal method, which then self-assembled into g-
C3N4 nanobers (CNFs) aer hydrothermal treatment. The
cobalt ions were absorbed by the CNF followed by cobalt
oxidation, until nally, they are converted into cobalt phos-
phide quantum dots by a low-temperature process. The CoP
QDs were monodispersed on the CNF, enhancing the electron–
hole pair separation efficiency and providing abundant active
sites for HER. So far, CoP/CNF has showed a HER rate of
2.42 mmol hꢀ1 in 15 vol% triethanolamine under visible light (l
$ 420 nm), which was 34.9 times higher than that of Pt/CNF.
The CoP/CNF heterojunction produced a large amount of gas
bubbles in a cuvette, which was observable by the naked eye
(Movie S1†). Meanwhile, the turnover number (TON) reached
519 with a turnover frequency (TOF) of 172.9 hꢀ1 for CoP QDs,
and the apparent quantum efficiency was 59.9% at 420 nm. So
far, to the best of our knowledge, CoP/CNF in our study has
shown one of the highest visible-light HER activities among all
g-C3N4-based photocatalysis systems and can even rival the best
visible light HER photocatalyst of the Ni2P QDs/CdS nanorods
heterogeneous nanojunction.16
that, the as-prepared product was collected by centrifugation
and washed with ethanol and deionized water. The obtained
composites with 1.0, 3.0, 5.0, 7.0 and 10.0 wt% Co3O4 QDs were
named as 1-Co3O4/CNF, 3-Co3O4/CNF, 5-Co3O4/CNF, 7-Co3O4/
CNF, and 10-Co3O4/CNF, respectively. Bare Co3O4 without g-
C3N4 nanober was prepared using the same method.
Synthesis of CoP QDs/g-C3N4 nanober (CoP/CNF) and bare
CoP
First, 20.0 mg Co3O4/CNF mixture with various amounts of
Co3O4 QDs and 100.0 mg of sodium hypophosphite (NaH2-
PO2$H2O, 105.99 g molꢀ1) were mixed together and ground to
a ne powder by using a mortar. Then, the mixture was calcined
at 280 C for 2 h with a heating speed of 2 C minꢀ1. The ob-
tained product was washed with deionized water and freeze
dried overnight to obtain 1-CoP/CNF, 3-CoP/CNF, 5-CoP/CNF, 7-
CoP/CNF, and 10-CoP/CNF, respectively. The same procedure
was used to prepare bare CoP.
ꢂ
ꢂ
Preparation of Pt/g-C3N4 nanober (Pt/CNF)
Experimental section
Various mass percents of Pt were in situ photodeposited on the
g-C3N4 nanober in 15 vol% triethanolamine (TEOA) including
a suitable amount of H2PtCl6 under room temperature and
a full arc light (l > 300 nm) for 2 h. Before irradiation, the mixed
solution was bubbled with high purity nitrogen for 45 min to
remove the air, ensuring that the reaction system was in an
inertial condition without oxygen reduction side reactions.
Finally, the appearance of a gray-black precipitate indicated that
the Pt particles were well deposited on the CNF and the
precipitate was renamed as 1 wt% Pt/CNF (1-Pt/CNF), 2 wt% Pt/
CNF (2-Pt/CNF), 3 wt% Pt/CNF (3-Pt/CNF), 4 wt% Pt/CNF (4-Pt/
CNF) and 5 wt% Pt/CNF (5-Pt/CNF).
The morphologies and chemical composition characteriza-
tion were performed using powder X-ray diffraction (XRD) with
Cu-Ka radiation (Rigaku, Smartlab), UV-vis diffuse reectance
spectrophotometer (DRS) (Varian, Cary 300), Fourier transform
infrared (FT-IR) spectroscopy (Nicolet 380, Thermo Fisher
Scientic), transmittance electron microscopy (TEM) (JEOL,
Chemicals
All reagent grade chemicals were purchased from Sinopharm
Chemical Reagent Co. Ltd and used without further purica-
tion. Ultrapure water was used in the whole experiment process
(Millipore-Q Century, 18.2 MU cm).
Synthesis of pristine g-C3N4
Pristine g-C3N4 was prepared by the polycondensation of
melamine (126.12 g molꢀ1, 6.0 g) from room temperature to
ꢂ
550 C under a N2 (99.999%) atmosphere at a ramping rate of
5 ꢂC minꢀ1 and kept for 4 h in a furnace, then cooled naturally
to room temperature and well ground by an agate mortar to
obtain pristine g-C3N4 powder.
Preparation of g-C3N4 nanober (CNF)
To prepare the g-C3N4 nanober, pristine g-C3N4 (0.92 g) was JEM-2100F), eld-emission scanning electron microscopy (FE-
dispersed in 60 mL distilled water, vigorously stirred overnight, SEM) (Hitachi, S-4800), and X-ray photoelectron spectroscopy
and then sonicated for 2 h. The bulky g-C3N4 particles were (XPS) with Al-Ka radiation (K-Alpha 1063, Thermo Fisher
removed from the ask bottom to put in suspension. Finally, Scientic). The photoluminescence (PL) spectra were recorded
the g-C3N4 suspension was added into an 80 mL volume stain- with a Hitachi F-7000 uorescence spectrophotometer. The
ꢂ
less steel vessel, and heated at 200 C for 6 h in an oven. Aer chemical element compositions were analyzed by EDS mapping
hydrothermal treatment, the solution was cooled down to room images captured on a Tecnai G2 F20 S-TWIN atomic resolution
temperature to obtain the product, which was named as CNF.
analytical
microscope.
Time-resolved
transient
This journal is © The Royal Society of Chemistry 2019
J. Mater. Chem. A, 2019, 7, 25908–25914 | 25909

View Article Online
Journal of Materials Chemistry A
Paper
photoluminescence (TRPL) spectroscopy was performed using nm). The average moles of evolved hydrogen in 3 h was deter-
an Hamamatsu universal streak camera C10910. Measurements mined at 7.26 ꢄ 10ꢀ3 mol in Fig. S11c.†
for the photocurrent Nyquist plots without bias potential were
The mass percentage of CoP QDs in the 7-CoP/CNF was
carried out with a CHI 660C electrochemical analyzer (CHI Inc., 6.32 wt% by EDS technology (Table S1†). Herein, the moles of
USA) in 0.5 M Na2SO4 aqueous solution with a three-electrode CoP QDs on 7-CoP/CNF was calculated by the equation:
conguration: FTO electrodes deposited with the samples
20 ꢄ 10ꢀ3 ꢄ 6.32%/(58.9 + 31.0) ¼ 1.4 ꢄ 10ꢀ5 mol.
acted as the photoanode, a platinum wire as the counter elec-
trode, and a saturated calomel electrode (SCE) as the reference
electrode. A 300 W xenon arc lamp with lighting wavelength
Thus, the TON and TOF were 519 and 172.9 hꢀ1
.
range of 320–780 nm (Perfect-light, PLS-SXE 300C, Beijing) was
used as the light source.
Results and discussion
Structure and component characterization
Photocatalytic test for water-splitting hydrogen evolution
reaction
Scanning electron microscopy (SEM) was used to investigate the
nanostructures and morphologies of CoP, CNF, and CoP/CNF.
CNF exhibited uniform ber-like morphology with ca. 100 nm
diameter, as shown in Fig. 1a. When Co3O4 was loaded on the
CNF with 1.0 wt% mass content, Co3O4 was well monodispersed
on the CNF without aggregation occurrence, as shown in
Fig. 1b, showing the 0D ultrane quantum dot structure. With
the Co3O4 mass content increased up to 3.0 wt%, the size also
got bigger, as clearly observed by SEM in Fig. 1c. Along with
further increasing the content up to 5.0, 7.0, and 10.0 wt%,
Co3O4 grew larger but did not grow together (Fig. 1d–f).
Comparatively, the bare Co3O4 without CNF as a host displayed
a solid micro-scale structure, growing together into a bulk
whole in Fig. S1.† The X-ray diffraction (XRD) pattern revealed
peaks at 18.9ꢂ, 31.2ꢂ, 36.9ꢂ, 38.5ꢂ, 44.9ꢂ, 55.8ꢂ, 59.3ꢂ, 65.3ꢂ,
74.2ꢂ, and 30.5ꢂ (Fig. S2†), which were indexed to the (111),
(220), (311), (222), (400), (422), (511), (440), (620), and (533)
facets of Co3O4 (JCPDS 42-1467),15 respectively. Besides, the
Co3O4 QDs on CNF possessed the same lattice structure as bare
Co3O4 (Fig. S3a†), showing CNF did not affect the crystal
structure of Co3O4. The X-ray energy dispersed spectrum (EDS)
in Fig. S3b† exhibited C, N, Co, and O elements in 7-Co3O4/CNF,
further conrming the existence of cobalt oxidation in the
Co3O4/CNF hybrids.
The photocatalytic HER was performed at room temperature. A
300 W xenon arc lamp was equipped with a 420 nm wavelength
cutoff lter and it had a light intensity of 120 mW cmꢀ2. The
photocatalyst sample (20.0 mg) was dispersed in TEOA solution
(15 vol%, 80 mL). H2 was analyzed using gas chromatography
˚
(Shimadzu, GC2010), equipped with a 5 A molecular sieve
column (Plot-Q column, 0.32 mm) and high purity nitrogen as
the carrier gas (99.999%).
Apparent quantum yield (AQE) calculation
The AQE was measured at 420 ꢃ 10 nm by a 300 W Xe lamp with
specic band-pass lters. The average irradiation intensity was
determined to be 2.39 mW cmꢀ2 and the irradiation area was
5.9 cm2. The amount of H2 generated in 5.0 h was 266.5 mmol.
Number of reacted electrons
AQE ¼
ꢄ 100%
Number of incident photons
Number of H2 molecules ꢄ 2
¼
ꢄ 100%
Number of incident photons
Number of H2 molecules ¼ 6.02 ꢄ 1023 ꢄ 266.5 ꢄ 10ꢀ6
¼
16.1 ꢄ 1019.
Number of incident photons ¼ El/hc ¼ (2.39 ꢄ 10ꢀ3 ꢄ 5.9 ꢄ
When processing the low-temperature phosphatization
treatment toward the bare Co3O4 under an argon atmosphere,
Co3O4 was well converted to CoP and showed a similar micro-
structure with the bare Co3O4 in Fig. S4.† Phosphorus-
3600 ꢄ 5.0 ꢄ 420 ꢄ 10ꢀ9)/(6.626 ꢄ 10ꢀ34 ꢄ 3 ꢄ 108) ¼ 5.36 ꢄ
1020
.
The calculated AQE at 420 ꢃ 10 nm was 59.9%.
Turnover number (TON) and turnover frequency (TOF)
calculations
The turnover number (TON) and turnover frequency (TOF) were
calculated by the following equations:
moles of evolved H2
TON ¼
moles of CoP QDs on photocatalyst
TON
TOF ¼
reaction time
The TON and TOF of CoP QDs on the 7-CoP/CNF nanober
Fig. 1 (a) SEM images of CNF, (b) 1-Co3O4/CNF, (c) 3-Co3O4/CNF, (d)
were measured under Xe lamp incident wavelength (l $ 420 5-Co3O4/CNF, (e) 7-Co3O4/CNF and (f) 10-Co3O4/CNF.
25910 | J. Mater. Chem. A, 2019, 7, 25908–25914
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
Journal of Materials Chemistry A
In the transmission electron microscopy (TEM) images,
numerous CoP QDs were observed and monodispersed on the
CNF without CoP QDs and CNF agglomeration (Fig. 3a–c). It
could also be clearly observed that the size of these CoP and the
integration of the 0D/1D CoP QDs/CNF nanohybrid were well
preserved. The diameter size of the CoP QDs remained at 6.8 ꢃ
0.2 nm, as seen in Fig. 3d. X-ray diffraction (XRD) patterns of 7-
CoP/CNF presented ve peaks at 31.6ꢂ, 36.3ꢂ, 46.3ꢂ, 48.2ꢂ, and
57.0ꢂ in Fig. 3e, corresponding to the (011), (111), (112), (211),
and (203) facets of CoP (JCPDS-29-0497),24 respectively, which
showed they had the same crystal structure as that of bare CoP
in Fig. S5.† In Fig. S6A,† the XRD records showed that the CoP
QDs hybridized with g-C3N4 nanober without affecting the
crystal structure of g-C3N4. In the FT-IR spectra in Fig. S6B,† the
810, 1200–1700, and 3200–3400 cmꢀ1 peaks were ascribed to the
tri-s-triazine ring out-of-plane bending, and N–C]N and NH2
stretching vibrations,25,26 respectively. The FT-IR spectrum
demonstrated that bare CoP had wide peaks at 1050 cmꢀ1, but
these were not found in the individual CNF. Comparatively, 7-
CoP/CNF had both characteristic peaks of CNF and CoP,
implying CoP was well loaded on the CNF. The energy dispersed
X-ray spectrum (EDS) disclosed the coexistence of Co and P in
the CNF (Fig. 3f), in a 1 : 1 ratio between Co and P and 6.32 wt%
CoP in 7-CoP/CNF (Table S1†), which was highly consistent with
the results from the atomic absorption spectroscopy (AAS)
analysis in Fig. S6C and S6D.† High-resolution transmission
electron microscopy (HR-TEM) was applied to determine the
crystal structures of CoP QDs. As shown in Fig. 3g and h, the
lattice fringes with the interplane distances of 0.306 and
0.197 nm were measured, corresponding to the (011) and (112)
facets of the CoP QDs,14 respectively. The corresponding
selected area electron diffraction (SAED) pattern in Fig. 3i
exhibited ve bright rings made up of discrete spots, which
Fig. 2 SEM images of: (a) phosphide CNF, (b) 1-CoP/CNF, (c) 3-CoP/
CNF, (d) 5-CoP/CNF, (e) 7-CoP/CNF and (f) 10-CoP/CNF.
containing particles were not observed on the CNF in Fig. 2a,
indicating the phosphatization treatment toward CNF did not
affect the CNF nanostructures. Aer the phosphatization on 1-,
3-, 5-, and 7-Co3O4/CNF, the Co3O4 QDs turned into CoP QDs
and grew larger than Co3O4, but without leading to agglomer-
ation problems, as seen from Fig. 2b–e. When 10-Co3O4/CNF
was suffering from low-temperature phosphatization, although
CoP was obtained, they were seriously stuck together on the
CNF, as shown in Fig. 2f. The SEM images demonstrated that
low-temperature phosphatization treatment could convert the
Co3O4 to CoP, but an excessive amount of Co3O4 on the CNF
could cause CoP connecting with adjacent particles, leading to
a bulk whole, causing the surface area and active sites to over-
lap, which is not benecial for HER.
Fig. 3 (a–c) Low-resolution TEM images of 7-CoP/CNF, (d) CoP QDs Fig. 4 (A) Photocatalytic HER rate of CNF hybridizing with various
diameter size distribution, (e) XRD pattern of 7-CoP/CNF and the CoP mass amounts of CoP QDs, (B) turnover number (TON) and turnover
standard card (JCPDS 29-0497), (f) energy dispersed X-ray spectrum frequency (TOF) of 7-CoP/CNF, (C) single-wavelength AQE versus UV-
(EDS) of 7-CoP/CNF, (g and h) high-resolution TEM images (HR-TEM) vis diffuse reflectance spectrum of 7-CoP/CNF in 15 vol% TEOA, (D)
of CoP QDs in 7-CoP/CNF and (i) selected area electron diffraction HER rate comparisons between 3-Pt/CNF and 7-CoP/CNF. Conditions
(SAED) of CoP QDs in 7-CoP/CNF.
(A, B, and D): 15 vol% TEOA and visible light (l $ 420 nm).
This journal is © The Royal Society of Chemistry 2019
J. Mater. Chem. A, 2019, 7, 25908–25914 | 25911

View Article Online
Journal of Materials Chemistry A
Paper
could be indexed to the (011), (201), (111), (211), and (301) facets the absorption edge was expanded to about 470 nm. Compar-
of the orthorhombic CoP QDs.27
atively, CNF demonstrated a red-shi of the absorption edge
aer hybridizing with CoP QDs. Moreover, 7-CoP/CNF had
narrower band gap (2.59 eV) than that of CNF (2.79 eV) (Fig. 5B).
Furthermore, the steady-state photoluminescence (PL) peak
intensity of 7-CoP/CNF remarkably decreased compared to CNF.
This measurement revealed that photogenerated electron–hole
pairs recombination in the 7-CoP/CNF had drastically
decreased compared to CNF (Fig. 5C and S8†), implying that the
electron–hole pairs recombination rate was efficiently sup-
pressed with the assistance of CoP QDs.4,5 Time-resolved pho-
toluminescence (TR-PL) spectroscopy in Fig. 5D proved the
photoexcited charge average emission lifetime was 6.39 ns for 7-
CoP/CNF, which was much shorter than that of CNF (8.28 ns)
(Table S3†). This result further conrmed the charge-transfer
pathway from CNF to CoP QDs in a nonradiative quenching
manner.21 Thereby, an efficient interface charge migration and
separation pathway were achieved in the 7-CoP/CNF. Further-
more, the electrochemical impedance spectroscopy (EIS) results
revealed that 7-CoP/CNF manifested a smaller charge-transfer
resistance (1.24 kU) than that of CNF (1.84 kU) and CoP (23.9
kU), as seen in Fig. S9a and Table S4,† conrming the lower
charge-transfer resistance in the 7-CoP/CNF heterostructure,
which permitted efficient separation of the photoinduced
charges.28 The identical ohmic resistance (0.16 kU) for CoP,
CNF, and 7-CoP/CNF was a consequence of the same concen-
tration of Na2SO4 electrolyte used during the experiments.
Moreover, the transient visible-light photocurrent indicated
that 7-CoP/CNF exhibited a higher photocurrent enhancement
than that of CNF and bare CoP (Fig. S9b†). All these observa-
tions demonstrated that CoP QDs could enhance the separation
and transportation of photoinduced charges, rendering a high
efficiency of photocatalysis for the HER.
Photocatalytic water-splitting hydrogen evolution
As shown in Fig. 4A, the photocatalytic HER performances of CNF
hybridizing with various mass amounts of CoP QDs in 15 vol%
triethanolamine (TEOA) under visible light (l $ 420 nm) were
identied. Bare CNF showed a negligible HER activity, while 7-
CoP/CNF exhibited the highest HER performance, which was
determined to be 2.42 mmol hꢀ1, 201.7 times higher than that of
the naked CNF. In the Pt/CNF hybrids, the optimal Pt content
loaded on CNF was 3 wt% and 3-Pt/CNF had the highest HER rate
of ca. 0.1 mmol hꢀ1 among Pt/CNF composites, as shown in
Fig. S7.† The 7-CoP/CNF produced a large number of gas bubbles
in a cuvette, observable by the naked eye (Movie S1†). The turn-
over number (TON) and turnover frequency (TOF) were 519 and
172.9 hꢀ1 for CoP QDs in the 7-CoP/CNF (Fig. 4B). The apparent
quantum efficiency (AQE) of 7-CoP/CNF reached as high as 59.9%
at 420 nm (Fig. 4C), presenting one of the best state-of-the-art g-
C3N4-based photocatalytic HER rates, higher than most semi-
conductor composites for photocatalytic HER (Table S2†).
Unexpectedly, the photocatalytic HER activity of 7-CoP/CNF was
34.9 times higher than that of 3-Pt/CNF in Fig. 4D, indicating CoP
may be a promising alternative to Pt-based catalysts in solar-to-
hydrogen fuel conversation.
Photoelectrochemical properties
As shown in Fig. 5A, the bare CoP exhibited full spectrum
absorption in the UV-visible light range due to its intrinsic black
color. CNF alone displayed a characteristic DRS spectrum and
Fig. 5 (A) UV-vis diffuse reflectance spectrophotometry (DRS) results
of CoP, CNF, and 7-CoP/CNF, (B) band gap of CNF and 7-CoP/CNF,
(C) steady-state photoluminescence (PL) spectra of bare CoP, CNF,
and 7-CoP/CNF under room temperature (inset: PL spectrum of bare Fig. 6 (A) X-ray photoelectron survey spectra (XPS) survey of CNF and
CoP and 7-CoP/CNF) and (D) time-resolved photoluminescence (TR- 7-CoP/CNF, (B) high-resolution N 1s XPS of CNF and 7-CoP/CNF, (C)
PL) spectrum through 375 nm light excited and collected at 480 nm high-resolution Co 2p XPS of 7-CoP/CNF and (D) high-resolution P 2p
under room temperature.
XPS of 7-CoP/CNF.
25912 | J. Mater. Chem. A, 2019, 7, 25908–25914
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
In Fig. 6A, the low-resolution X-ray photoelectron spectros-
Journal of Materials Chemistry A
copy (XPS) survey spectrum displayed C, N, and O elements in
the CNF, while in comparison, both Co and P were detected in
the 7-CoP/CNF, showing the presence of CoP aer the phos-
phatization of 7-Co3O4/CNF. In Fig. S10,† the high-resolution C
1s XPS at 284.5 and 288.0 eV are for the standard reference
carbon and sp2-bonded C in N]C–N,29 respectively. Three pairs
of N 1s peaks at 398.5, 399.1, and 400.7 eV in Fig. 6B correspond
to C]N–C, N–(C)3, and N–H of CNF,30 respectively. The slight
shi of the C]N–C and N–(C)3 peaks of 7-CoP/CNF was likely
due to their lone pair of electrons, which can serve as Co metal-
coordination sites. The high-resolution Co 2p XPS peak could
be deconvoluted into two peaks at 789.9 and 783.0 eV, which
correspond to Co 2p3/2 and Co 2p1/2,
31 respectively (Fig. 6C). In
Fig. 7 (a–c) TEM images of 7-CoP/CNF, (d) CoP QDs diameter size
distribution, (e) EDS of 7-CoP/CNF and (f) HR-TEM of CoP QDs.
Fig. 6D, the P 2p peak of 7-CoP/CNF, was deconvoluted to three
peaks at 129.7, 130.6, and 133.6 eV, corresponding to the P 2p2/
3, P/2p1/2, and P–O species,32 respectively. The XPS proved that
CoP QDs were well hybridized with CNF and displayed a solid
interaction with CNF, which was benecial to photoexcited
charges direct separation and photocatalysis for the HER.
95.6% (Fig. S11†). Based on the TEM images in Fig. 7, the
nanoheterojunction structure and composition of 7-CoP/CNF
remained stable aer long-time cyclic HER. The CNF retained
the uniform nanober morphology and CoP kept the quantum
dot shape without growing into bulky microspheres (Fig. 7a and
S12a†). CoP QDs were clearly detected in the HR-TEM (Fig. 7b–
d) without obvious aggregation issues aer the cycling experi-
ments. Based on EDS analysis, the Co content kept stable aer
long-term cycling test compared to fresh 7-CoP/CNF (Fig. 7e and
Table S5†). HR-TEM showed the CoP QDs special lattice fringe
d ¼ 0.233, 0.290, and 0.245 nm, corresponding to (222), (220),
and (311) facets in Fig. S12b† and 7f. The high-resolution XPS of
7-CoP/CNF indicated the CoP QDs were stable without obviously
being oxidized before and aer the cycling experiment, as seen
in Fig. S13.† In summary, the nanostructure and chemical
composition of 7-CoP/CNF were extremely stable aer the long-
term cycling test.
Photocatalytic hydrogen evolution mechanism
Based on the results above, a possible photocatalytic hydrogen
evolution mechanism on CoP/CNF is proposed in Scheme 2.
Under visible light irradiation, protons were absorbed by the g-
C3N4 nanober and the electrons in the valence band of the g-
C3N4 nanober were excited, and migrated to its conduction
band, and were then transferred to CoP quantum dots for
proton reduction hydrogen generation. The generated holes in
the valence band oxidized the sacricial agent TEOA. During the
photocatalysis hydrogen evolution process, the CoP quantum
dots in CoP/CNF not only helped separate the electron–hole
pairs and suppress electron–hole recombination, but also acted
as a cocatalyst and provided abundant sites for proton reduc-
tion hydrogen evolution.
Conclusions
Photocatalytic performance and nanostructure stability
In conclusion, a visible light-driven system using CoP QDs as
a HER cocatalyst to form an integrated nanostructure with g-
C3N4 nanober for photocatalytic HER was exploited. The
reaction components in the system consisted of catalysts made
of earth-abundant elements only, which made this system quite
intriguing for the conversion of solar energy to chemical fuels.
CoP QDs could highly promote electron–hole pair separation
and the subsequent reduction of water to generate hydrogen.
The photocatalytic activity of CoP/CNF was extremely high, with
impressive HER performance, which was one of the highest
among g-C3N4-based hybrids reported to date. Furthermore, the
idea of loading transition-metal phosphides (TMPs) QDs on
metal-free g-C3N4 nanostructures could be an efficient design
that can guide and benet future research regarding the
development of photocatalytic hydrogen generation.
The nanostructure stability of a catalyst is considered as one of
the most indispensable factors when used in practical appli-
cations. Aer nine cycles, the HER performance of 7-CoP/CNF
was highly stable and its catalytic activity remained as high as
Conflicts of interest
Scheme 2 Photocatalytic hydrogen generation mechanism on CoP/
CNF under visible light.
The authors declare no conict of interests.
This journal is © The Royal Society of Chemistry 2019
J. Mater. Chem. A, 2019, 7, 25908–25914 | 25913

View Article Online
Journal of Materials Chemistry A
Paper
15 P. Jiang, Q. Liu, C. Ge, W. Cui, Z. Pu, A. M. Asiri and X. Sun, J.
Mater. Chem. A, 2014, 2, 14634–14640.
Acknowledgements
16 Z. Sun, H. Zheng, J. Li and P. Du, Energy Environ. Sci., 2015, 8,
2668–2676.
17 W. Liu, L. Cao, W. Cheng, Y. Cao, X. Liu, W. Zhang, X. Mou,
L. Jin, X. Zheng, W. Che, Q. Liu, T. Yao and S. Wei, Angew.
Chem., Int. Ed., 2017, 56, 9312–9317.
18 A. Li, W. Zhu, C. Li, T. Wang and J. Gong, Chem. Soc. Rev.,
2019, 48, 1874–1907.
19 S. Hua, D. Qu, L. An, W. Jiang, Y. Wen, X. Wang and Z. Sun,
Appl. Catal., B, 2019, 240, 253–261.
The authors appreciate the support from the China Scholarship
Council for Y. Zeng and Y. Xia, and Hunan Provincial Innova-
tion Foundation for Postgraduate (CX2017B139 and
CX2017B141). Meanwhile, thanks Dr Jinming Luo of Georgia
Institute of Technology, Dr Qingmei Chen of Boston College
and Mr Joey Yourg of University of Illinois at Urbana-Cham-
paign for grammar and expression corrections. The views and
ideas expressed herein are solely those of the authors and do
not represent the ideas of the funding agencies in any form.
20 C. Yang, Z. Xue, J. Qin, M. Sawangphruk, S. Rajendran,
X. Zhang and R. Liu, J. Phys. Chem. C, 2019, 123, 4795–4804.
21 Y. Zeng, C. Liu, L. Wang, S. Zhang, Y. Ding, Y. Xu, Y. Liu and
S. Luo, J. Mater. Chem. A, 2016, 4, 19003–19010.
22 Y. Deng, B. Chi, X. Tian, Z. Cui, E. Liu, Q. Jia, W. Fan,
G. Wang, D. Dang and M. Li, J. Mater. Chem. A, 2019, 7,
5020–5030.
23 J. Fu, Q. Xu, J. Low, C. Jiang and J. Yu, Appl. Catal., B, 2019,
243, 556–565.
24 L. Hu, Y. Liu, G. Zeng, G. Chen, J. Wan, Y. Zeng, L. Wang,
H. Wu, P. Xu, C. Zhang, M. Cheng and T. Hu,
Chemosphere, 2017, 184, 1071–1079.
References
1 T. Hisatomi and K. Domen, Nat. Catal., 2019, 2, 387–399.
2 M. Sonnberger, Nat. Energy, 2019, 4, 265–266.
3 Y. Zeng, H. Li, J. Luo, J. Yuan, L. Wang, C. Liu, Y. Xia, M. Liu,
S. Luo and T. Cai, Appl. Catal., B, 2019, 249, 275–281.
4 Y. Zeng, X. Liu, C. Liu, L. Wang, Y. Xia, S. Zhang, S. Luo and
Y. Pei, Appl. Catal., B, 2018, 224, 1–9.
5 W. Cui, N. Cheng, Q. Liu, C. Ge, A. M. Asiri and X. Sun, ACS
Catal., 2014, 4, 2658–2661.
6 J. Tian, Q. Liu, Y. Liang, Z. Xing, A. M. Asiri and X. Sun, ACS
Appl. Mater. Interfaces, 2014, 6, 20579–20584.
7 J. Liu, Y. Liu, D. Xu, Y. Zhu, W. Peng, Y. Li, F. Zhang and
X. Fan, Appl. Catal., B, 2019, 241, 89–94.
8 Z. Sun, H. Zheng, J. Li and P. Du, Energy Environ. Sci., 2015, 8,
2668–2676.
25 D. Wu, S. Hu, H. Xue, X. Hou, H. Du, G. Xu and Y. Yuan, J.
Mater. Chem. A, 2019, 7, 20223–20228.
26 J. Ran, T. Pan, Y. Wu, C. Chu, P. Cui, P. Zhang, X. Ai, C.-F. Fu,
Z. Yang and T. Xu, Angew. Chem., Int. Ed., 2019, 58, 1–8.
27 P. Jiang, Q. Liu, C. Ge, W. Cui, Z. Pu, A. M. Asiri and X. Sun, J.
Mater. Chem. A, 2014, 2, 14634–14640.
9 Z. Pu, Q. Liu, C. Tang, A. M. Asiri and X. Sun, Nanoscale,
2014, 6, 11031–11034.
28 M. A. Hossain, D. J. Burritt and M. Fujita, Abiotic Stress
Response in Plants, 2016, pp. 323–375.
10 P. Jiang, Q. Liu, Y. Liang, J. Tian, A. M. Asiri and X. Sun,
Angew. Chem., 2014, 126, 13069–13073.
11 J. Tian, Q. Liu, N. Cheng, A. M. Asiri and X. Sun, Angew.
Chem., 2014, 126, 9731–9735.
12 Z. Xing, Q. Liu, A. M. Asiri and X. Sun, Adv. Mater., 2014, 26,
5702–5707.
13 Z. Xing, Q. Liu, A. M. Asiri and X. Sun, ACS Catal., 2014, 5,
145–149.
29 Y.-J. Yuan, Z. Shen, S. Wua, Y. Su, L. Pei, Z. Jia, M. Ding,
W. Bai, Y. Chena, Z.-T. Yu and Z. Zou, Appl. Catal., B, 2019,
246, 120–128.
30 Y. Ding, Y. Tang, L. Yang, Y. Zeng, J. Yuan, T. Liu, S. Zhang,
C. Liu and S. Luo, J. Mater. Chem. A, 2016, 4, 14307–14315.
31 T. Liu, Q. Liu, A. M. Asiri, Y. Luo and X. Sun, Chem. Commun.,
2015, 51, 16683–16686.
32 T. Liu, P. Li, N. Yao, G. Cheng, S. Chen, W. Luo and Y. Yin,
Angew. Chem., Int. Ed., 2019, 58, 4679–4684.
14 Q. Liu, J. Tian, W. Cui, P. Jiang, N. Cheng, A. M. Asiri and
X. Sun, Angew. Chem., Int. Ed., 2014, 53, 6710–6714.
25914 | J. Mater. Chem. A, 2019, 7, 25908–25914
This journal is © The Royal Society of Chemistry 2019