Photocatalytic hydrogen evolution from biomass (glucose solution) on Au/CdS nanorods with Au3+ self-reduction

Article abstract of DOI:10.1016/j.jssc.2020.121495

Glucose, as one of the most familiar biomass, can be utilized to produce hydrogen and fine chemicals. This work reports the renewable hydrogen production from glucose over Au/CdS-nanorods (Au/CdS-NRs) photocatalyst. For Au/CdS-NRs catalyst, it shows a superior H₂ evolution rate of 90.0 ?μmol ?g^?1 ?h^?1 under visible light irradiation. Moreover, the H₂ generation from the glucose solution on Au/CdS-NRs prepared by twice solvothermal method is much higher than Au colloid and photoreduction methods. Prepared with twice solvothermal method, Au nanoparticles are obtained from Au³⁺ ions on the CdS surface, as Au³⁺ can be reduced by photogenerated electrons. The separated hole can oxidize glucose to produce protons, and protons can be reduced by photogenerated electrons into H₂. Meanwhile, the Au/CdS-NRs can also be used to produce H₂ from other biomasses, for instance sucrose, soluble starch, celluflor, pectin and waste paper scraps. The probable mechanism is proposed and discussed in the photocatalytic process from glucose reforming. This work indicates that the process of self-reducing Au³⁺ to Au can drastically increase the photocatalytic efficiency of H₂ evolution from biomass, by promoting the separation and transportation of photogenerated charge carriers.

Full text of DOI:10.1016/j.jssc.2020.121495

Journal Pre-proof
Photocatalytic hydrogen evolution from biomass (glucose solution) on Au/CdS
3+
nanorods with Au self-reduction
Xulin Wang, Xiuzhen Zheng, Huijuan Han, You Fan, Sujuan Zhang, Sugang Meng,
Shifu Chen
PII:
S0022-4596(20)30325-X
DOI:
https://doi.org/10.1016/j.jssc.2020.121495
YJSSC 121495
Reference:
To appear in: Journal of Solid State Chemistry
Received Date: 19 April 2020
Revised Date: 26 May 2020
Accepted Date: 3 June 2020
Please cite this article as: X. Wang, X. Zheng, H. Han, Y. Fan, S. Zhang, S. Meng, S. Chen,
3+
Photocatalytic hydrogen evolution from biomass (glucose solution) on Au/CdS nanorods with Au self-
reduction, Journal of Solid State Chemistry (2020), doi: https://doi.org/10.1016/j.jssc.2020.121495.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
©
2020 Published by Elsevier Inc.

Credit Author Statement
Xulin Wang: Experiment, Methodology, Software
Xiuzhen Zheng: Data curation, Writing- Original draft preparation.
Huijuan Han: Experiment
You Fan, Sujuan Zhang, Sugang Meng: Methodology, Reviewing
Shifu Chen: Writing- Reviewing and Editing

Graphical Abstract

Photocatalytic hydrogen evolution from biomass (glucose solution) on
3
+
Au/CdS nanorods with Au self-reduction
a
a, b
a
a
a
a
Xulin Wang , Xiuzhen Zheng* , Huijuan Han , You Fan , Sujuan Zhang , Sugang Meng , Shifu
a
Chen*
a
Key Lab of Clean Energy and Green Circulation, Huaibei Normal University, Anhui Huaibei,
235000, People's Republic of China
b
3
Shanghai Key Laboratory of Atmospheric Particle Pollution and Prevention (LAP ), Fudan
University, Shanghai, 200433, P. R. China
*Corresponding Authors, Tel: +86-561-3803225, Fax: +86-561-3803225. E-mail:
zhengxz@chnu.com, chshifu@chnu.edu.cn.
1

Abstract:
Glucose, as one of the most familiar biomass, can be utilized to produce hydrogen and fine
chemicals. This work reports the renewable hydrogen production from glucose over
Au/CdS-nanorods (Au/CdS-NRs) photocatalyst. For Au/CdS-NRs catalyst, it shows a superior H2
−
1
−1
evolution rate of 90.0 μmol g h under visible light irradiation. Moreover, the H generation from
2
the glucose solution on Au/CdS-NRs prepared by twice solvothermal method is much higher than Au
colloid and photoreduction methods. In the twice solvothermal method, Au nanoparticles are
3
+
3+
obtained from Au ions in glucose solution, as Au can be reduced by photogenerated electrons.
The separated hole can oxidize glucose to produce protons, and protons can be reduced by
photogenerated electrons into H . Meanwhile, the Au/CdS-NRs also exhibit some produce H form
2
2
other biomasses, for instance sucrose, soluble starch, celluflor, pectin and waste paper scraps. The
probable mechanism is proposed and discussed in the photocatalytic process from glucose reforming.
3
+
This work indicates that the process of self-reducing Au to Au can drastically increase the
photocatalytic efficiency of H evolution from biomass, which promotes the separation and
2
transportation of photogenerated charge carriers.
Keywords: biomass; glucose; Au/CdS-nanorods; self-reduction; visible light; hydrogen evolution.
2

1. Introduction
Biomass, a most versatile renewable resource, can be used for hydrogen production, which is a
promising approach to sustainably generate fuels and feedstock chemicals [1, 2]. Glucose as a
polyhydroxy aldehyde, is the most widely distributed and important monosaccharide in nature. It has
been studied as model substrates of biomass to produce H , which especially plays a significant role
2
in the biological field and is the main energy-generating substance of organisms. Tremendous efforts
have been devoted to producing fuels directly from glucose by photocatalysis [3-5], which as a green,
safe and promising technology in the biomass reforming, does not require high temperature and
pressure [6-8].
Solar-to-hydrogen from glucose can solve energy supply and storage problems, which has
attracted much interest of the researchers [9-11]. Semiconductor nanomaterials are widely used as
the photocatalysts, due to their tunable optical and electronic properties. Among them, metal
sulphides based semiconductors as the excellent photocatalysts for H evolution are invented to
2
address the energy issue [12, 13]. Especially, the band gap of CdS is close to the optimum value for
effective solar energy conversion, and has a high absorption coefficient in the visible range of solar
spectrum [14]. However, the photocatalytic efficiency of CdS is generally limited by the slow
transfer of photoexcited charge carriers and subsequently high recombination rates of charges
[
15-17]. Many strategies have been developed to solve this issue by designing CdS-based
photocatalysts [18], such as loading cocatalyst, synthesizing sulfide solid solution (Mn Cd S [19],
x
1-x
Co Cd S [12] and Zn Cd S [20]), and recombining with other semiconductors (TiO [21],
0
.2
0.8
0.5
0.5
2
BiVO [22]). Among them, the combination of cocatalyst with CdS is an effective method to
4
improve the photocatalytic performance of hydrogen production, including metal (Pt [23], Au [24],
and Co [25]), metal oxide (NiO [26], Co O [27]), transition metal sulfide (MoS [28, 29], NiS [30])
3
4
2
and phosphide (Ni P [31], MoP [32]). It is known that noble metals have been widely used as
2
cocatalysts [33-39], which can significantly accelerate the separation and transportation of
photoinduced charge carriers. However, the effect of noble metal ion reported so far is very limited,
3

3
+
θ
3+
since they are easy to be reduced. For instance, Au is easy to be reduced to Au [E (Au /Au) =
+
θ
+
1
.40 eV], which is more favorable compared with H reduction to H [E (H /H ) = 0 eV] [36]. Thus,
2 2
3
+
3+
for the Au modified CdS photocatalyst, when Au are reduced by photoexcited electrons
transferred from CdS), the holes are left in the valance band of CdS, which can oxidize glucose to
(
generate protons [34]. This transfer route can effectively separate the photogenerated electrons and
holes. Moreover, the self-reduced Au nanoparticles would form schottky junctions with CdS, which
can further accelerate the charge transfer. Therefore, self-reduction of noble metal ion can drastically
increase the photocatalytic efficiency of H evolution from biomass by promoting the separation and
2
transportation of photogenerated electron–hole pairs.
It is a difficult process for photocatalytic splitting of glucose to H , as glucose is not easy to
2
+
dissociate H . Herein, we develop a twice-solvothermal process to prepare Au/CdS-NRs composite
3
+
with Au doped on the surface of CdS nanorods (CdS-NRs). 1D CdS-NRs is selected as the
substrate due to its unique properties, such as short radial distance for separation and transportation
of photoexcited electrons-holes, which is favorable for electron-directed conduction. The optimally
prepared Au/CdS-NRs exhibit high activity and stability for photocatalytic H evolution in glucose,
2
verified by other biomasses (sucrose, soluble starch, celluflor, pectin and waste paper scraps) under
visible light (λ > 400 nm) irradiation. Especially, this approach can be used for other
Au/semiconductor to improve the efficiency of photocatalytic H evolution. The probable
2
mechanism is proposed by the photoelectrochemical method and time-resolved photoluminescence
(
TR-PL), indicating that Au is an excellent co-catalyst for photocatalytic H evolution.
2
2
2
. Experimental
.1. Materials
Cd(NO ) • 4H O, CS(NH ) , ethylenediamine, HAuCl solution and glucose were purchased from
3
2
2
2 2
4
Aladdin Industrial Inc.. All the reagents were analytical pure reagents and deionized water was used
in the whole work.
2.1. Fabrication of CdS-NRs and Au/CdS-NRs.
CdS-NRs were synthesized via a solvothermal method [37]. Firstly, 4.7 g Cd(NO ) • 4H O, 3.5 g
3
2
2
CS(NH ) (thiourea) and 60 mL ethylenediamine were added into a 100 mL Teflon liner to form a
2
2
4

yellow solution after stirring 0.5 h. Then the liner is encased into a stainless-steel autoclave and
heated at 180 ℃ for 48 hrs. After the reaction, the autoclave was cooled down and the products are
washed with ethanol and deionized water for several times. Finally, the product was dried in vacuum
at 60 ℃ for 12 hrs to obtain CdS nanorods.
Au/CdS was further prepared by solvothermal method. Firstly, 1.0 g CdS (obtained from the above
preparation) and 50 mL ethylenediamine was uniformly dispersed into Teflon liner. Then, 632, 1263,
1
895 and 2527 μL HAuCl solution (0.1 g/ml) and 2 mL ethanol were respectively added into CdS
4
suspension, and the weight ratio of added Au with CdS are about 1%, 2%, 3% and 4%. The
corresponding Au loading measured by ICP-MS is 0.66%, 1.71%, 2.57% and 3.85%, respectively.
The mixture was also heated at 150 ℃ for 12 hrs. After cooled naturally to room temperature, the
resultant products were washed with water/ethanol for several times and finally dried in vacuum at
80 ℃ for 12 hrs.
2.3. Characterization.
The morphology, phase composition, optical properties, valence status, photoelectrochemical
properties and lifetime (τ) of photoexcited electrons of the samples were respectively investigated by
scanning electron and transmission electron microscopy (SEM and TEM), X-ray diffraction (XRD),
UV-vis diffuse reflectance spectrometer (DRS), X-ray photoelectron spectroscopy (XPS)
measurements, electrochemical workstation and time-resolved PL spectra (the details are shown in
Supplementary Material).
2.4. Evaluation of photocatalytic properties.
The photocatalytic hydrogen production performance of the prepared Au/CdS was tested on a
LabSolar II reaction system (Perfect Light Co.). 100 mg CdS is dispersed in a reactor containing
00-mL aqueous solution (5 g/L glucose), and then connected to the hydrogen production reaction
1
system. Prior to irradiation, the suspension was stirred for 30 minutes to achieve
adsorption-desorption equilibrium. A 300 W Xe lamp (PLS-SEX 300, Perfect Light Co., Beijing)
equipped with a UV cut filter (λ > 400 nm) was used as the visible light source. After turning on the
lamp, Au deposited on the CdS-NRs is used as the photocatalyst for hydrogen evolution from
glucose. During the test, the temperature of the solution controlled by the water circulation device
was kept at 10 °C. During the reaction, the generated H was measured by an on-line gas
2
5

chromatograph (GC7900, TianMei, Shanghai).
3
3
. Results and discussion
.1 Characterization of Au/CdS nanorods
Fig. 1. The SEM images of CdS-NRs (a) and 3% Au/CdS-NRs (b) with element mapping (c-e), TEM
(f) and HR-TEM (g) of 3% Au/CdS.
Fig. 1 shows the SEM and TEM microstructural characterization of CdS-NRs and 3%
Au/CdS-NRs samples. The rod-like structure is clearly observed with a length 0.5-3 μm and a
diameter 50-100 nm (Fig. 1a). Moreover, Au particles with large size (more than 100 nm, Fig. 1b and
S1) are formed by small nanoparticles aggregated on CdS surface, which are easily differentiated
from CdS rods in the SEM and TEM images. From Fig. S2, it can be found that Au particles are
small and homogeneously distributed on the surface of CdS-NRs (1% Au/CdS-NRs), which become
6

reunion from the Au amount with 2%. This may be due to that ethylenediamine (en) is used as the
3
+
sole solvent without water, which may form [Au(en) ] complex on the surface of CdS, and then
3
assemble to form big particles [40-42]. Additionally, the elemental mapping images (Fig. 1c-e)
confirm that the modified CdS is composed of Au, Cd, and S elements, of which Cd and S elements
are homogeneously distributed along the rods. The TEM and HRTEM observation (Fig. 1f and 1g)
further shows the rod-shape structure of CdS, and the lattice fringes of 0.245 nm and 0.235 nm are
distinct. They are corresponding to the typical d-spacing value for (102) plane of CdS nanorods and
(111) plane of Au nanoparticles [43], and the intimate contact of interplanar spacing is formed
between Au and CdS.
Fig. 2. XRD patterns of CdS-NRs, Au/CdS-NRs with different Au loading amounts and standard
hexagonal CdS (JPCDS: 41-1049).
Fig. 2 shows the XRD patterns of pure CdS and Au/CdS samples synthesized with different Au
loading amounts. For pure CdS, the diffraction peaks are respectively ascribed to the corresponding
diffraction planes, indexed well to the hexagonal CdS phase structures (JCPDS: 41-1049) [44]. The
obvious diffraction peaks of 2θ = 24.9°, 26.6°, 28.3° and 43.8° correspond to (100), (002), (101) and
(110) crystal faces of hexagonal phase CdS. As for Au/CdS composites, besides the typical
diffraction peaks of CdS, the characteristic peaks of Au (38.2° and 64.6°, ascribed to JCPDS No.
04-0784) are also found (labeled with ♦) [45], and the intensity are increased with the increase of Au
7

content.
The optical properties of the prepared Au/CdS samples are investigated by UV-vis DRS technique.
The light absorption of these samples are shown similar absorption band edge positions (Fig. 3),
which can be attributed to the intrinsic band gap absorption. The absorption edge of CdS locates at
about 520 nm, and the corresponding band-gap energy is calculated to be 2.42 eV. After Au deposited
on CdS rods, the absorption edge of Au/CdS samples showed a slight red-shift, and the band-gap
energy are slightly reduced, centered at about 2.4 eV. Moreover, an obvious absorption enhancement
at around 740 nm is observed in the visible light region, due to surface plasmon resonance absorption
produced from large size of Au particle, which is consistent with the earlier studies [46]. Therefore,
the doposition of Au on the CdS surface is an effective way to expand and enhance the absorbption
of CdS in visible light region.
Fig. 3. UV-vis DRS spectra and band-gap energy (inset) of CdS-NRs and Au/CdS-NRs with different
3
+
Au loading contents.
8

Fig. 4. (a) XPS survey spectra of 3% Au/CdS and the high-resolution XPS spectra of (b) Cd 3d, (c) S
2p, and (d) Au 4f.
To further investigate the modification of Au on CdS surface, the constitution and the valence
status of the prepared CdS and Au-modified CdS are further detected via XPS technique. For 3%
Au/CdS, Au, Cd and S element signals can be observed in the survey spectrum (Fig. 4a). The carbon
and oxygen peak is attributed to the absorbed gaseous molecules (O and CO ) on the catalyst
2
2
surface. The high-resolution XPS spectrum of Cd 3d, S 2p, and Au 4f are respectively shown in Fig.
4
4
b, c, and d. It indicates that the binding energies (BEs) of Cd 3d5/2 and Cd 3d located at 404.6 and
3/2
11.4 eV with a divided energy of 6.8 eV, and the BEs of S 2p3/2 and S 2p1/2 can be deconvoluted
0
3+
into two peaks centered at 161.0 and 162.2 eV [12]. As for Au 4f spectra (Fig. 4d), Au and Au
peaks are simultaneously detected. The binding energies of Au 4f7/2 (82.8 eV) are deviated 3.7 eV
0
from Au 4f5/2 eV (86.5 eV), which are attributed to the metallic Au species [44]. The Au 4f and Au
7/2
3
+
4
f binding energy of Au is to be 84.1 and 87.5 eV, which is near standard deviation of Au metal
5/2
3
+
0
(
3.35 eV) [47]. From the XPS analysis, it can be observed that Au is partially reduced to Au in the
0
synthetic process, with ethanol as reducing agent.Why Au species contains two valence status (Au
3
+
3+
and Au ) in the 3% Au/CdS? There are two reasons for the Au existence: (1) the weak reducing
9

ability of ethanol. Au cannot be fully reduced and aggregated in the high temperature and pressure.
2) High concentration of reaction solution (HAuCl ). As ethylenediamine (en) is used as the sole
(
4
3+
solvent without water in Au/CdS synthesis, which may form [Au(en) ] complex attached on the
3
CdS surface.
To consider the interaction of Au and CdS, the shift of BEs is investigated. The shift of Au 4f in
0
Au/CdS is 1.2 eV, which is lower than that of the Au metal (Au , the peaks of 4f and 4f5/2 are at
7
/2
8
4.0 eV and 87.7 eV), while the shifts of Cd 3d5/2 and Cd 3d increase compared with that of CdS
3/2
(403.8 and 410.6 eV, Fig. S3). These shifts could be caused by the interaction between Au and CdS.
3
+
The transfer of electrons from CdS to the more electropositive Au may explain this binding energy
shift.
3
.2 Photocatalytic performance of H evolution reaction
2
The photocatalytic performances of H evolution on the Au/CdS samples have been investigated
2
from glucose and other biomasses under visible light irradiation (Fig. 5). As seen from Fig. 5a, there
is no H production detected without light or on the pristine CdS catalysts, which means that CdS
2
cannot produce H evolution from glucose by photocatalytic process. However, Au/CdS catalysts
2
with different Au contents produce much higher amount of H , suggesting that the amount of Au
2
loading has a significant effect on the photocatalytic activity in the Au/CdS composites. The H yield
2
is firstly increased and then decreased with the increase of the Au loading amount. Little or much
amount of Au loading covered on the surface of CdS may lead to the high electron-hole
recombination, and 3 % is the appropriate amount of cocatalyst for CdS loading in the system.
−
1
−1
Correspondingly, its H production rate is about 90 μmol g h , and the amount reaches 350 μmol
2
after reaction for 4 hrs in glucose solution. Moreover, the CO production is also detected (Fig. 5b).
2
The ratio between hydrogen evolution and CO production is about 1:1, and the increasing tendency
2
is also close to that of H . This means that H and CO are simultaneously generated from glucose in
2
2
2
the photocatalytic process.
1
0

Fig. 5. The amount of (a) H and (b) CO produced by glucose, (c) the amount of H produced by
2
2
2
other biomass (sucrose, soluble starch, celluflor and pectin) on 3% Au/CdS with different reaction
time, and (d) the amount of H after reaction prolong to 17 hrs on 3% Au/CdS under visible light
2
irradiation.
In order to make clear the reason of biomass hydrogen evolution and develop practical method to
improve hydrogen evolution rate, sucrose, soluble starch, celluflor and pectin are also used as
research biomasses to investigate the H production. As shown in Fig. 5c, sucrose, soluble starch,
2
celluflor and pectin are also able to produce H on 3% Au/CdS, but the yields are much lower than
2
that of glucose at the same reaction time. This may be due to that glucose as a polyhydroxy aldehyde
(small molecule), is more easily dehydrogenated than macromolecular biomass. Meanwhile, an
expansion of producing H from waste paper scraps was tested (Fig. S4). After 1 month of natural
2
degradation of the paper dust and filter of the liquid, the hydrogen production activity test (5 hrs
−
1
−1
under light irradiation) was carried out. The hydrogen production rate is 27.6 μmol g h , and the
1
1

-1
total hydrogen production is 138 μmol g . The waste paper scraps can generate hydrogen by
photocatalysis, which realizes the reuse of waste and provides a good example of full utilization from
world resources.
Fig. 6. The XRD patterns (a), high-resolution XPS spetra of Au (b) and TEM images (c and d) of 3%
Au-CdS after photocatalytic reaction.
Besides, the stability of Au/CdS photocatalysts was also evaluated. As can be seen from Fig. 5d,
the time experiment is prolonged for 17 hrs on 3% Au/CdS in the glucose solution. The H evolution
2
rate is gradually decrease after 9 hrs, indicating that 3% Au/CdS composite performs a certain degree
of stability in biomass hydrogen production. Although its XRD and XPS data (Fig. 6a and 6b)
3
+
display no significant change after 4 hrs reaction, the Au concentration is decreased as the reaction
3
+
0
proceeding. This means that the Au is self-reduced to Au , which is tested by photoluminescence
3
+
method (Fig. S5). [48] Moreover, the reduction process was proposed. The Au precursor (Au ) and
3+
ethylenediamine (en) may form a strong ligand compound [Au(en) ] complex attached on the CdS
3
3
+
3+
surface. In the photocatalytic process, [Au(en) ] will releases the Au slowly, which makes the
3
1
2

3
+
Au long-timely reduced by photogenerated electrons in the reaction system.
Meanwhile, the Au morphology is changed from large particles to small one. This may be due to
that H is produced on the aggregated Au surface, changing its morphology. In addition, the CdS
2
3
+
nanorods are acted as the seed materials for Au heterogeneous nucleation in the photocatalytic
process. The photoinduced electrons generated in the CdS can transfer to the surface, and the
3
+
accumulated electrons on the surface of the CdS nanorods can reduce Au ions to Au seeds. The Au
3
+
seeds on the CdS surface can play an important role in electron storage, and thus, the Au ions in the
solvent are adsorbed and reduced to nanoparticles (< 10 nm). The separated holes are trapped by
+
glucose, which as the holes’ scavenger produces H and CO . This charge transfer route can
2
3
+
effectively separate the photogenerated electrons and holes, with Au self-reduced to Au
nanoparticles, which can be confirmed by TEM and HR-TEM images (Fig. 6c and 6d). It is clearly
found that many Au nanoparticles are formed on the surface, and schottky junctions will be
established with CdS, which is favorable for photogenerated electrons transferring from the CdS to
+
Au. Then, H can obtain electrons to generate H .
2
To further confirm the effect of self-reduction, 3% Au/CdS are also prepared by other methods
shown in Fig. 7a (Au colloid and photoreduction methods, the detailed prepared methods are shown
in Supplementary Material), and Au nanoparticles are formed on the surface of CdS-NRs (Fig. 7c
and 7d). However, no H is observed from the photocatalytic process (Fig. 7b). Measured with
2
ICP-MS, the Au content are 2.57%, 1.52% and 2.42% for 3% Au/CdS prepared with twice
solvothermal, Au colloid and photoreduction, respectively. The Au loading amount with colloid
method is much lower than that with twice hydrothermal and photoreduction. Combined with the
3
+
above analysis, the PL intensity at 380 nm is clearly decreased with the increase of the Au ions. As
the intensity of Au/CdS prepared with twice solvothermal is much lower than that of Au/CdS
3
+
prepared with Au colloid and photoreduction (Fig. S6), the Au concentration of Au/CdS prepared
by twice solvothermal method is much higher than that of Au/CdS prepared by other methods. Thus,
3
+
self-reduction of Au and schottky junctions on the Au/CdS-NRs sample can improve the
photocatalytic activity of H evolution by separating and effectively transferring photogenerated
2
electrons.
1
3

Fig. 7. The XRD pattern (a) and H production (b) of 3% Au/CdS-NRs prepared with different
2
methods, and TEM images of 3% Au/CdS prepared by (c) Au colloid and (d) photoreduction
methods.
3
.4 Photocatalytic H production mechanism
2
The photogenerated charge behaviors are characterized by photoelectrochemical method and
photoluminescence technique. Instantaneous photocurrent response and electrochemical impedance
EIS) testing has proven to be the effective means of studying the efficiency of photogenerated
(
electron-hole pair separation [49, 50]. Fig. 8a exhibits the transient photocurrent response of
Au/CdS-NRs and CdS-NRs, in which 3% Au/CdS-NRs shows much higher photocurrent than that of
other samples. The presence of Au as a co-catalyst does facilitate the separation of photogenerated
electron-hole pairs. Due to the excellent electrical conductivity of Au, the transfer process is
accelerated, so the recombination of photoexcited charge carriers is suppressed to some extent. The
effect of Au on the separation efficiency of photoexcited carriers are also investigated by
1
4

electrochemical impedance (EIS) testing. The radius of the arc in the Fig. 8b represents the charge
transfer process that occurs on the electrode surface. The smaller the radius is, the faster transfer of
interface charge is, with the separation for the photogenerated electron-hole pairs largely promoted.
The result of EIS Nyquist plots follows the order: CdS < 1% Au/CdS < 2% Au/CdS < 4% Au/CdS <
3
% Au/CdS, which is consistent with the results of photocurrent and photocatalytic H evolution.
2
Fig. 8. The transient photocurrent response (a) and electrochemical impedance spectroscopy (b) of
Au/CdS samples under visible light.
In addition to investigate the separation of photoexcited electron-hole pair, another key factor
affecting the efficiency of hydrogen production is the band edge potential of the photocatalyst, which
determines whether the hydrogen production can proceed smoothly from thermodynamics. In order
to investigate the potential effect, the energy band structures of the prepared CdS and series of
Au/CdS were measured by M-S curve. As shown in Fig. 9, a positive slope of the linear region
indicates that the CdS sample is a n-type semiconductor [19]. According to the intercept of the
X-axis tangent, the flat band potential (E ) of CdS can be calculated to be −0.30 V (vs. NHE). Since
fb
the E of the n-type semiconductor is lower than 0.1 V of E generally, the conduction band edge
cb
fb
(
E ) of CdS is about −0.40 V. Thus, the E of 1% Au/CdS, 2% Au/CdS, 3% Au/CdS and 4%
cb cb
Au/CdS estimated from the M-S plots is ca. −0.43, −0.45, −0.53, and −0.58 V, respectively.
Obviously, the E position of 3% Au/CdS is more negative 1.87 and 0.53 V (vs. NHE, pH 7) than
cb
3
+
+
that of Au and H reduction potential, thermodynamically ensuring that Au nanoparticles and H
2
can successfully be reduced.
1
5

1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
CdS
1
2
3
4
% Au/CdS
% Au/CdS
% Au/CdS
% Au/CdS
-0.8 -0.6 -0.4 -0.2
0.0
0.2
0.4
Potential (V vs. NHE)
Fig. 9. The Mott-Scotty plot of Au/CdS with different Au amount.
Additionally, the recombination rate of the photoinduced electron-hole pairs of the sample is
further analyzed by PL spectroscopy and time-resolved photoluminescence decay technology.
Generally, higher PL intensities exhibit higher recombination rates of photoinduced electron and
holes. As shown in Fig. 10a, the PL spectrum of the Au/CdS-NRs sample with an excitation
wavelength of 400 nm has an emission peak in the range of 450-700 nm. The strength of CdS-NRs is
significantly higher than that of Au/CdS-NRs composites, indicating that the photoexcited carrier
recombination rate of CdS composites is higher than that of Au/CdS. In order to verify the above
speculations, time-resolved photoluminescence decay technology is used to further study the
separation efficiency of photo-generated carriers (Fig. 10b). The calculated lifetime of the 3%
Au/CdS-NRs is 22.9 ns, which is longer than that of CdS-NRs (14.2 ns), confirming the charge
transfer from CdS to Au. The above results are consistent with the photocatalytic activity of the
Au/CdS photocatalyst and EIS spectroscopy.
1
6

Fig. 10. (a) Photoluminescence (PL) and (b) time-resolved PL spectra of CdS-NRs and a series of
Au/CdS-NRs with different Au amount
As shown in Fig. 11a, the work function of CdS (4.10 eV) is lower than that of Au (5.35 eV) [51,
52]. Along with the contact of Au and CdS, the Fermi levels of the Au nanoparticles and CdS align,
resulting in the band bending of CdS and the formation of Schottky barrier. Under visible light
irradiation, the photoexcited electrons flow from CdS-NRs with a lower work function into the Au
with a higher work function (Fig. 11b), due to that the Fermi level of Au nanoparticles is much lower
than the bended conduction band of CdS. Moreover, the as-formed Schottky barrier can promote the
transfer of electrons from the CdS to the Au, and leave holes in the VB of ZnO, decreasing the
recombination of photogenerated electrons and holes (Fig. 11c).
Based on these analyses, the reaction mechanism is discussed as following. The reaction is
initiated by photoexcitation of Au/CdS particles, which results in the formation of electron–hole pairs.
3
+
+
0
Au and H dissociated from glucose can trap the photogenerated electron to produce Au and
hydrogen [37], and the holes left on the valence band of CdS-NRs can oxidize glucose (Fig. 11d)
−
θ
[
(
13]. Theoretically, the accumulated holes (2.02 eV) cannot oxidize H O or OH to produce •OH [E
2
−
OH /•OH) = 2.40 eV] due to the less negative potential. However, the active species of •OH is still
detected in the photocatalytic reaction by ESR spin-trapping technique (Fig. S7), which may come
–
from O • not the hole, as the •OH amount is not too much. The holes and •OH can oxidize the
2
θ
+
glucose to form CO and H [E (C O H /CO , H ) = 0.014 V], and some intermediates are
2
2
6
6
12
2
1
7

produced by dehydrogenation and decarboxylation reaction (measured by LC-MS in Fig. S8), which
may be arabinose (C H O ), erythrose (C H O ), 5-hydroxymethyl furfural (C H O ), levulinic acid
5
10
5
4
8
4
6
6
3
(
C H O ) and 3-formylpropionicacid (C H O ) [53]. Thus, in the glucose reforming to produce H ,
5 8 3 4 6 3 2
+
the photoinduced holes are firstly dehydrogenation (H ) and then decarboxylation (CO ) to form
2
intermediates, which as the fine chemicals may be possess higher economic value than glucose.
Fig. 11. Energy level lineup diagrams for Au and CdS (a) before and (b) after interfacing. (C)
Schematic band diagram illustrating the charge transfer under UV-visible light irradiation (Evac and
EF denote vacuum level and Fermi level, respectively). (d) The schematic illustration for the
mechanism of photocatalytic hydrogen evolution over Au/CdS in glucose solution.
4. Conclusion
A series of rod-like Au/CdS composites with varied content of Au were successfully prepared by
twice solvothermal method. The binary Au/CdS composites show higher photocatalytic
performances than pure CdS, and exhibit the optimal activity when the composite consists of 3% Au.
1
8

−
1
−1
The highest H evolution rate is 90.0 μmol g h in glucose solution under visible light irradiation.
2
Significantly, the activity of 3% Au/CdS photocatalyst remained stable after a continuous test (9 hrs).
Moreover, some biomass (sucrose, soluble starch, celluflor, pectin and waste paper scraps) are also
3
+
able to produce H . The self-reduction effect of Au and schottky junctions in 3% Au/CdS
2
composite are the main reasons in improving the H evolution reaction from biomass, which can
2
efficiently promote the separation and transportation of the photogenerated holes and electrons. This
work suggests the possibility of hydrogen generation from cheap and abundant biomass and open a
pathway for constructing excellent heterojunction photocatalysts.
Acknowledgement
This work was supported by the Natural Science Foundation of China (NSFC, grant Nos.
51772118, 51972134 and 21707027), the Natural Science Foundation of Educational Committee of
Anhui Province (grant No. KJ2019A0601 and KJ2018A0387), Opening Project of Shanghai Key
3
Laboratory of Atmospheric Particle Pollution and Prevention (LAP ) (Grant No. FDLAP19007),
the Project of Anhui Province for Excellent Young Talents in Universities (Grant No. gxyq2019029),
and Graduate Innovation Foundation of Huaibei Normal University (ycx201901003).
References:
[
[
1] M.F. Kuehnel, E. Reisner, Angew. Chem. Int. Edit., 57 (2018) 3290-3296.
2] L.I. Granone, F. Sieland, N. Zheng, R. Dillert, D.W. Bahnemann, Green Chem., 20 (2018)
1
[
1
[
169-1192.
3] N. Shen, F. Zhang, X.N. Song, Y.S. Wang, R.J. Zeng, Int. J. Hydrogen Energ., 38 (2013)
1241-11247.
4] M. Cappelletti, G. Bucchi, J.D. Mendes, A. Alberini, S. Fedi, L. Bertin, D. Frascari, J. Chem.
Technol. Biot., 87 (2012) 1291-1301.
5] V. Vaiano, G. Iervolino, G. Sarno, D. Sannino, L. Rizzo, J.J.M. Mesa, M.C. Hidalgo, J.A. Navio,
Oil Gas Sci. Technol., 70 (2015) 891-902.
[
[
[
(
[
[
[
[
6] Y. Chen, B. Zhai, Y. Liang, Y. Li, J. Li, J. Solid State Chem., 274 (2019) 32-39.
7] D. Li, J. Sun, T. Shen, H. Song, J. Liu, C. Wang, X. Wang, R. Zhao, J. Solid State Chem., 286
2020) 121301.
8] A. Singh, D.P. Dutta, A. Ballal, A.K. Tyagi, M.H. Fulekar, Mater. Res. Bull., 51 (2014) 447-454.
9] L. Sun, Y.C. Yin, F. Wang, W.X. Su, L.X. Zhang, Dalton Trans., 47 (2018) 4315-4324.
10] L. Zhang, X. Hao, Q. Jian, Z. Jin, J. Solid State Chem., 274 (2019) 286-294.
11] L.Z. Zhang, J.W. Shi, M.C. Liu, D.W. Jing, L.J. Guo, Chem. Commun., 50 (2014) 192-194.
1
9

[
[
12] Q. Deng, T. Miao, Z. Wang, Y. Xu, X. Fu, Chem. Eng. J., 378 (2019) 122139.
13] Y. Li, L.L. Wang, T. Cai, S.Q. Zhang, Y.T. Liu, Y.Z. Song, X.R. Dong, L. Hu, Chem. Eng. J.,
321 (2017) 366-374.
[
[
14] S. Meng, X. Ning, S. Chang, X. Fu, X. Ye, S. Chen, J. Catal., 357 (2018) 247-256.
15] S. Rengaraj, S. Venkataraj, S.H. Jee, Y. Kim, C.W. Tai, E. Repo, A. Koistinen, A. Ferancova, M.
Sillanpaa, Langmuir, 27 (2011) 352-358.
[
[
[
16] Z. Sun, H. Chen, L. Zhang, D. Lu, P. Du, J. Mater. Chem. A, 4 (2016) 13289-13295.
17] Y. Cao, G. Wang, Q. Ma, Z. Jin, Dalton Trans., (2019).
18] Q. Wang, J. Li, Y. Bai, J. Lian, H. Huang, Z. Li, Z. Lei, W. Shangguan, Green Chem., 16 (2014)
2
[
4
728-2735.
19] H. Li, Z. Wang, Y. He, S. Meng, Y. Xu, S. Chen, X. Fu, J. Colloid Interface Sci., 535 (2019)
69-480.
[
[
20] R. Sun, J. Song, H. Zhao, X. Li, Mater. Charact., 144 (2018) 57-65.
21] Q.R. Liu, X.Y. Tan, S.B. Wang, F. Ma, H. Znad, Z.F. Shen, L.H. Liu, S.M. Liu, Environ.
Sci-Nano, 6 (2019) 3170-3170.
[
[
22] L. Zou, H.R. Wang, X. Wang, ACS Sustain. Chem. Eng., 5 (2017) 303-309.
23] J. Fang, J.J. Gu, Q.L. Liu, W. Zhang, H.L. Su, D. Zhang, ACS Appl. Mater. Inter., 10 (2018)
1
[
9649-19655.
24] S. Oros-Ruiz, A. Hernandez-Gordillo, C. Garcia-Mendoza, A.A. Rodriguez-Rodriguez, R.
Gomez, J. Chem. Technol. Biot., 91 (2016) 2205-2210.
25] J.A. Nasir, M. Hafeez, M. Arshad, N.Z. Ali, I.F. Teixeira, I. McPherson, Zia-ur-Rehman, M.A.
Khan, ChemsusChem, 11 (2018) 2587-2592.
[
[
[
[
(
[
[
(
[
26] F.Y. Chen, L.J. Zhang, X.W. Wang, R.B. Zhang, Appl. Surf. Sci., 422 (2017) 962-969.
27] O. Yehezkeli, D.R.B. de Oliveira, J.N. Cha, Small, 11 (2015) 668-674.
28] S.W. Zhang, H.C. Yang, H.H. Gao, R.Y. Cao, J.Z. Huang, X.J. Xu, ACS Appl. Mater. Inter., 9
2017) 23635-23646.
29] K. Chang, X. Hai, J. Ye, Adv. Energ. Mater., 6 (2016) 1502555.
30] J.L. Yuan, J.Q. Wen, Y.M. Zhong, X. Li, Y.P. Fang, S.S. Zhang, W. Liu, J. Mater. Chem. A, 3
2015) 18244-18255.
31] W.L. Zhen, X.F. Ning, B.J. Yang, Y.Q. Wu, Z. Li, G.X. Lu, Appl. Catal. B-Environ., 221 (2018)
243-257.
[
(
[
[
32] S. Duan, S. Zhang, S. Chang, S. Meng, Y. Fan, X. Zheng, S. Chen, Inte. J. Hydrogen Energ., 44
2019) 21803-21820.
33] J. Kim, K. Yong, A facile, J. Nanopart Res., 14 (2012).
34] X.L. Song, Y.M. Liu, Y. Zheng, K. Ding, S.J. Nie, P. Yang, Phys. Chem. Chem. Phys., 18 (2016)
4577-4584.
[
[
35] H. Cao, Z. Liu, T. Liu, S. Duo, L. Huang, S. Yi, L. Cai, Mater. Charact., 160 (2020) 110125.
36] V. Kumaravel, M.D. Imam, A. Badreldin, R.K. Chava, J.Y. Do, M. Kang, A. Abdel-Wahab,
Catalysts, 9 (2019)276.
37] X. Fu, L. Zhang, L. Liu, H. Li, S. Meng, X. Ye, S. Chen, J. Mater. Chem. A, 5 (2017)
5287-15293.
[
1
[
[
[
38] D.P. Dutta, A. Singh, A.K. Tyagi, J. Environ. Chem. Eng., 2 (2014) 2177-2187.
39] M.H. Fulekar, A. Singh, D.P. Dutta, M. Roy, A. Ballal, A.K. Tyagi, RSC Adv., 4 (2014) 10097.
40] A. Zaleska-Medynska, M. Marchelek, M. Diak, E. Grabowska, Adv. Colloid Interfac., 229
2
0

(
[
[
2016) 80-107.
41] M.C. Daniel, D. Astruc, Chem. Rev., 104 (2004) 293-346.
42] L. Li, Q.K. Kong, Y. Zhang, C.M. Dong, S.G. Ge, J.H. Yu, J. Mater. Chem. B, 3 (2015)
2
[
764-2769.
43] S.Y. Moon, H.C. Song, E.H. Gwag, I.I. Nedrygailov, C. Lee, J.J. Kim, W.H. Doh, J.Y. Park,
Nanoscale, 10 (2018) 22180-22188.
[
[
44] J. Zhang, S.Z. Qiao, L. Qi, J. Yu, Phys. Chem. Chem. Phys., 15 (2013) 12088-12094.
45] M. Ahmad, Y.Y. Shi, A. Nisar, H.Y. Sun, W.C. Shen, M. Wei, J. Zhu, J. Mater. Chem., 21 (2011)
7723-7729.
[
[
46] J.S. Cai, J.Y. Huang, Y.K. Lai, J. Mater. Chem. A, 5 (2017) 16412-16421.
47] T. Ueda, K. Ueda, K. Ito, K. Ogasawara, H. Kanetaka, T. Mokudai, Y. Niwano, T. Narushima, J.
Biomed Mater. Res. A, 107 (2019) 991-1000.
48] S. Obregon, A. Vazquez, M.A. Ruiz-Gomez, V. Rodriguez-Gonzalez, Applied Surface Science,
88 (2019) 205-212.
[
4
[
[
(
[
49] H. Zhang, R. Zong, Y. Zhu, J. Phys. Chem. C, 113 (2009) 4605-4611.
50] X. Ding, D. Xiao, L. Ji, D. Jin, K. Dai, Z.X. Yang, S.Y. Wang, H. Chen, Catal. Sci. Technol., 8
2018) 3484-3492.
51] C. Wang, Z.Y. Hu, H. Zhao, W.B. Yu, S.J. Wu, J. Liu, L.H. Chen, Y. Li, B.L. Su, J. Colloid Interf.
Sci., 521 (2018) 1-10.
[
[
52] D. Yamashita, A. Ishizaki, Appl. Surf. Sci., 363 (2016) 240-244.
53] J. Zou, G. Zhang, X. Xu, Appl. Catal. A-Gen., 563 (2018) 73-79.
2
1

Highlights
ò Au/CdS nanorods was prepared by twice solvothermal method.
3
+
ò Au can be reduced by photogenerated electrons in glucose solution.
3
+
ò Self-reduction of Au can significantly improve the charge separation.
ò A photocatalytic mechanism was proposed and confirmed by several techniques.

Declaration of Interest Statement
We develop a twice-solvothermal process to prepare Au/CdS-NRs composite with
3
+
Au doped CdS nanorods (CdS-NRs). The optimally prepared Au/CdS-NRs exhibit
high activity and stability for photocatalytic H evolution in glucose, verified by other
2
biomasses (sucrose, soluble starch, celluflor, pectin and waste paper scraps) under
3
+
visible light (λ > 400 nm) irradiation. The self-reduction effect of Au and schottky
junctions in 3% Au/CdS composite are the main reasons in improving the H2
evolution reaction from biomass, which can efficiently promote the separation and
transportation of the photogenerated holes and electrons. It is expected that our work
could give inspiration to generate H from cheap and abundant biomass and open a
2
pathway for constructing excellent heterojunction photocatalysts.