A reaction environment modulation strategy was employed to promote the H2 production over plasmonic Au/semiconductor composites. It is shown that the fast consumption of the holes in plasmonic Au nanoparticles by methanol in alkaline reaction environment remarkably increases H2 generation rate under visible light. The photocatalytic reaction is mainly driven by the interband transition of plasmonic Au nanoparticles, and the apparent quantum efficiency of plasmon-assisted H2 production at pH 14 reaches 6% at 420 nm. The reaction environment control provides a simple and effective way for the highly efficient solar fuel production from biomass reforming through plasmonic photocatalysis in future.

Photocatalysis is a green technology for solar-to-chemical energy conversion. In general, the photocatalytic performance can be improved through two approaches. The first one is to improve the materials by means of energy band engineering, morphology control, construction of heterojunctions, modification of cocatalysts, etc.1–5 Another alternative is the modulation of the reaction environment to enhance the photocatalytic performance, by which the surface properties of photocatalysts or the reaction pathways can be controlled.6–8 Photocatalytic H2 production is an important subject in solar fuel production. Of particular interest in the recent years is to use the alkaline environment to realize highly efficient H2 production in aqueous system containing alcohol (as sacrificial reagent) over conventional oxide or sulfide semiconductors under visible light.6,7 This reaction involves the photooxidation of alcohols by the valence band holes and the photoreduction of protons by conduction band electrons. The enhanced photocatalytic H2 production in alkaline environment is generally considered to benefit from the non-Nernstian dependence of the band edges shift of semiconductors on the pH of reaction system (i.e., slope ≠ − 0.059 V/pH).6,7

Plasmonic photocatalysis is the cutting-edge research field of visible-light photocatalysis and has found extensive applications in H2 production, environmental remediation, and organic transformation in the past few years.9–15 The redox ability of plasmonic photocatalysts originates from the surface plasmon resonance (SPR).13 Au nanoparticles (NPs) are the most popular and widely used visible-light harvesting plasmonic photocatalysts up to now. The lifetime of the hot electrons in the SPR process of Au NPs is very short (in the time scale of 10−3 ns).16,17 Therefore, plasmonic Au NPs are often used in conjunction with a semiconductor (such as TiO2). The ultrafast hot electron injection from Au NPs to semiconductor can effectively extend the lifetime of the hot electrons (into the time scale of 100 ns) to favor the photocatalytic reactions.16,18,19 On the basis of the current understanding of electron transfer mechanism in plasmonic photocatalysis,19,20 the redox process of H2 production through plasmonic photocatalysis can be schematically shown in Figure 1. Although plasmonic photocatalysis provides an alternative approach for visible-light H2 production, the performance is still far less satisfactory when compared with the achievements of conventional semiconductor photocatalysis.21,22 This is basically because the lifetimes of hot electrons through SPR excitation in Au NPs or Au/TiO2 system are still much shorter than that of the exited charge-carriers in semiconductors (e.g., TiO2>103 ns),23 implying that the recombination of e/h+ in plasmonic Au NPs is much more frequent than semiconductors. Therefore, to suppress the recombination of hot electrons and holes in Au NPs is crucial to enhance the efficiency of plasmonic photocatalysts.

FIG. 1.

Schematic process of H2 production through plasmonic photocatalysis.

FIG. 1.

Schematic process of H2 production through plasmonic photocatalysis.

Close modal

Herein, the alkaline environment is employed to enhance the H2 production over plasmonic Au/TiO2 photocatalyst. This work was inspired by the growing awareness of the vital role of an alkaline environment for noble metal Au as well as Pt and Pd in showing their excellent electrocatalytic alcohol oxidation performance.24–30 Among these noble metal electrocatalysts, Au is a lucky one having photocatalytic performance because of its distinctive SPR effect. It is reasonable that the mechanism of electrocatalytic redox reaction over Au can provide inspiring knowledge to Au-based plasmonic photocatalysis. This is essentially determined by the identical physical nature in both processes, that is, the former process is driven by the externally applied bias and the latter is driven by the photovoltaic effect of exited Au SPR. In this work, plasmonic photocatalysis successfully drew the lessons from electrocatalytic methanol oxidation mechanism to improve the plasmon-assisted H2 production over Au/TiO2 photocatalysts. It is proposed that, in alkaline reaction system, the fast consumption of the holes in plasmonic Au NPs by methanol effectively suppresses the recombination of hot electrons and holes, leading to enhanced H2 production and the high apparent quantum efficiency (AQE).

Two types of classic Ti-based photocatalytic materials,TiO2 (commercial P25) and SrTiO3 (STO), were used as supports of Au NPs in this work. The large difference of their conduction band positions (nearly 1 eV)31,32 can help us estimate to what extent the band edge positions of a semiconductor support will influence the performance of plasmonic photocatalyst (as discussed below). The plasmonic Au/P25 and Au/STO composite photocatalysts were prepared by conventional deposition-precipitation method.9–11,33 The detailed preparation process of STO and the deposition-precipitation method refer to our previous report.10 The loading amounts of Au for different samples were determined by inductively coupled plasma optical emission spectroscopy (ICP-OES) method. Thus obtained Au NPs have relatively uniform particle size distribution (3-5 nm, see TEM images in Figure S134 and related references).10,33 The UV-Vis spectra of Au/P25 and Au/STO showed that the SPR of Au NPs induced a wide photoabsorption in visible light range (Figure S2)34 with the peak centered at ∼550 nm. The performance of H2 production of 1.2 wt. % Au/P25 loaded with Pt cocatalyst at various pH values was measured under visible light (Figure 2). When the pH value is lower than 13, only a trace amount of H2 evolution was found over Au/P25 photocatalysts. However, the H2 production rate was remarkably enhanced when the pH value reached 14. In 5M NaOH solution, the H2 production rate achieved more than 200 μmol h−1. Similar results were also obtained over Au/STO samples (Figure S4).34 Activities of the samples with different Au loading amounts on P25 and STO were also measured at pH 14 (Figure S5).34 The optimized Au loading amount was ∼0.5 wt. %, indicating that effective H2 production can be realized in alkaline reaction environment by sensitizing TiO2 or STO with tiny amount of plasmonic Au NPs. However, the reason for this optimized loading amount is not clear at the present stage. The photocatalysts also showed stable photocatalytic activity in alkaline environment. H2 production almost kept a linear increase after overnight reaction over Au/P25 sample (Figure S6a).34 Both TiO2 and STO maintained stable in alkaline environment during the photocatalytic reaction (Figure S6b).34 Control experiment using 0.9 wt. % Au/3.6 wt. % Pt/SiO2 did not lead to H2 evolution at pH 14, implying that a semiconductor support is important to realize the separation of hot electrons to drive the photocatalytic reactions. In addition, control experiment using 0.25 wt. % Au/P25 without Pt cocatalyst in pH 7 and 14 just produced a trace amount of H2 within 6 h. Such H2 production amount is lower than the detection limit of gas chromatograph unless the photocatalytic reaction can be further prolonged. As far as we know, the H2 production activity measured in alkaline environment by this work is obviously higher than the recent reports on plasmon-assisted H2 production (Table S1).34 The AQE of H2 production over 0.42 wt. % Au/P25 photocatalysts was measured (see the test details in the supplementary material)34 in the visible light range at pH 14 (Figure 3(a)). It was observed that the AQE increased linearly from 0% to 6% with the decrease of wavelength from 730 to 420 nm, indicating that the photocatalytic reaction in alkaline environment can be initiated by almost the whole range of visible light photons. AQE measurement over 0.49 wt. % Au/STO showed a similar trend (Figure S7).34 It can be easily speculated that the level of AQE will be further improved at higher pH values than 14.

FIG. 2.

(a) H2 production over 1.2 wt. % Au/P25 at varied pH values under visible light irradiation (300 W Xenon lamp with L42 cutoff filter, λ > 400 nm). (b) H2 production rates in (a). The photocatalytic H2 production was carried out in a Pyrex glass reaction cell containing 50 ml CH3OH and 220 ml H2O. Twenty-five milligram of photocatalyst was used. Five weight percentage Pt cocatalyst was loaded by photodeposition (Figure S3).34 The pH of solution was adjusted by NaOH.

FIG. 2.

(a) H2 production over 1.2 wt. % Au/P25 at varied pH values under visible light irradiation (300 W Xenon lamp with L42 cutoff filter, λ > 400 nm). (b) H2 production rates in (a). The photocatalytic H2 production was carried out in a Pyrex glass reaction cell containing 50 ml CH3OH and 220 ml H2O. Twenty-five milligram of photocatalyst was used. Five weight percentage Pt cocatalyst was loaded by photodeposition (Figure S3).34 The pH of solution was adjusted by NaOH.

Close modal
FIG. 3.

(a) The apparent quantum efficiency of H2 evolution over 0.42 wt. % Au/P25 photocatalysts in the visible light range. Reaction was carried out under same condition as Figure 2 with the pH fixed 14. (b) The schematic absorption spectrum of Au NPs (black line). The spectrum consists of the absorptions from both intraband and interband excitations.

FIG. 3.

(a) The apparent quantum efficiency of H2 evolution over 0.42 wt. % Au/P25 photocatalysts in the visible light range. Reaction was carried out under same condition as Figure 2 with the pH fixed 14. (b) The schematic absorption spectrum of Au NPs (black line). The spectrum consists of the absorptions from both intraband and interband excitations.

Close modal

The hot electrons in a plasmonic nanometal can be generated through both the intraband (between the Fermi level and sp conduction band) and interband transitions (between d band and sp conduction band) during the SPR decay.16,19 The contributions of intraband and interband transitions to a typical UV-Vis spectrum of ordinary Au NPs without special nanostructures are schematically shown in Figure 3(b).11,35–37 It is noteworthy that the AQE profile is in a perfect agreement with photoabsorption through interband transition of Au NPs, which is similar to our previous report.10 This result can be possibly explained from two aspects. On one hand, the oxidation ability of Au originates from the holes generated by SPR excitation of Au NPs.9,16,38 Quantitative investigation using nanosecond pulse-laser has evidenced that the ionization of Au NPs is mainly contributed by the interband excitation.37 Therefore, the interband excitation can be directly correlated with the photooxidation ability of Au NPs to methanol. On the other hand, seeing from the energy band alignment, the d band is lying lower than Fermi level of gold.16 This means that the potential of the holes formed by interband transition should be more positive than that formed by intraband transitions and the former should have stronger photooxidation ability to methanol. These facts indicate that interband transition plays more dominative role in the photooxidation process than intraband transition, which will be finally reflected on the feature of AQE.

STO support has a much more negative conduction band level than P25 TiO2 and will render higher reduction potential to the separated hot electrons from Au NPs for proton reduction. This indeed contributed subtle H2 production to Au/STO in comparison with only trace amount of H2 evolution over Au/P25 at low pH range (Figure 2(b) and Figure S4b34). However, this enhancement is almost negligible when compared with the enhanced H2 production by increasing the pH value for each sample (Figure 2(b) and Figure S4b34). In fact, Au/P25 showed higher activities than Au/STO at high pH range. Moreover, in theory, the upward bending of the surface band of semiconductor in high pH value will increase the injection barrier of hot electrons and play adverse effect on H2 evolution performance. Therefore, the mechanism of the enhanced H2 production over semiconductor photocatalysts (non-Nernstian shift of band edges)6 is not completely applicable to explain how the alkaline environment enhanced the H2 production over plasmonic Au/semiconductor photocatalysts. We speculate that the alkaline environment may exert direct influence on the surface redox reaction. Because the proton reduction will be generally suppressed in concentrated alkaline environment, it is very likely that the alkaline environment enhanced the methanol photooxidation by the holes in the plasmonic Au NPs. Actually, the electrocatalytic oxidation of alcohols (methanol, ethanol, glycerol, etc.) over Au can only readily occur under alkaline environment, whereas under acidic and neutral conditions, the reaction becomes significantly sluggish.24–28 Koper and coworkers proposed that the alkaline environment dominated the main process of alcohol oxidation over Au electrode which involved two deprotonation steps, including a “base-catalyzed” step to produce reactive alkoxide at high pH values and a further “gold catalyzed” step for alkoxide oxidation.24 The oxidation behavior of these alcohols generally varies in a highly nonlinear fashion with pH, showing a highly consistent trend with the variation of photocatalytic H2 production rate with pH in this work.

The (photo) electrochemical analysis is a powerful tool to simulate the microscopic electron transfer mechanism and provide the information related to the surface chemical reaction pathway in a photocatalytic reaction.6,7,39,40 Photoelectrochemical measurements in this work further provided conclusive evidences for the above mentioned interpretations. The working electrodes were fabricated by spin-coating the well dispersed Au/P25 or pure P25 suspension on indium tin oxide (ITO) substrates followed by a heat treatment at 400 °C (see the details in the supplementary material).34 Under the dark condition, the Au NPs on the photoanode just served as electrocatalysts and there was a sudden increase of the dark current when the pH reached 14 (Figure 4(a)). This result showed a consistent conclusion with the previous electrocatalytic studies, confirming that alkaline environment can effectively enhance the methanol oxidation over Au.24–28 An obvious inflection point at pH 14 was also observed in the photoresponse of Au/P25 film (Figure 4(b)), clearly showing that, under alkaline environment, enhanced photooxidation of methanol was achieved by the SPR excitation of Au NPs to realize the solar-to-chemical energy conversion. Pure TiO2 is incapable of effectively driving the methanol oxidation electrocatalytically or photocatalytically under visible light, leading to negligible current responses in control experiments (Figures 4(a) and 4(b)). For Au/TiO2 composite, the injected hot electron to TiO2 will decay back to the Au NPs just after ∼1.5 ns to recombine with a hole if no donor can be timely consumed on Au NPs.16,18,19 The fast photooxidation of methanol enhanced consumption of the holes in plasmonic Au NPs and suppressed the recombination of e/h+ pairs, resulting in effective regeneration of hot electrons for photocatalytic H2 production.

FIG. 4.

Photoelectrochemical methanol oxidation of Au/P25 and P25 electrodes at varied pH values. (a) Dark current density and (b) photoresponse measured at 1.1 V vs. RHE (reversible hydrogen electrode). The value of photoresponse here is obtained by subtracting the dark current from the total photocurrent (schematically shown in Figure S8).34 The electrolyte is same as photocatalytic reaction solution and the pH was adjusted by NaOH. A 500 W Xenon lamp equipped with L42 cutoff filter was used as the visible light source.

FIG. 4.

Photoelectrochemical methanol oxidation of Au/P25 and P25 electrodes at varied pH values. (a) Dark current density and (b) photoresponse measured at 1.1 V vs. RHE (reversible hydrogen electrode). The value of photoresponse here is obtained by subtracting the dark current from the total photocurrent (schematically shown in Figure S8).34 The electrolyte is same as photocatalytic reaction solution and the pH was adjusted by NaOH. A 500 W Xenon lamp equipped with L42 cutoff filter was used as the visible light source.

Close modal

In summary, alkaline environment modulation remarkably boosts H2 production over plasmonic Au/semiconductor photocatalysts and achieves exceptional performance. This strategy could be universally applicable to the generation of hydrogen from aqueous system containing a variety of alcohols. The fast photooxidation process effectively accelerates the consumption of holes in Au NPs and improves the AQE of plasmonic photocatalysis. In view of the growing importance of solar-to-chemical energy conversion, this is in particular attractive for the solar fuel production from photoreforming of polyhydroxy biomass over Au-based plasmonic photocatalysts under visible light. This work also hints that electrocatalytic reactions over gold are transplantable for their applications in plasmonic photocatalysis.

This work received financial support from the World Premier International Research Center Initiative (WPI Initiative) on Materials Nanoarchitectonics (MANA), MEXT, Japan, and National Basic Research Program of China (973 Program, 2014CB239301).

1.
H.
Tong
,
S.
Ouyang
,
Y.
Bi
,
N.
Umezawa
,
M.
Oshikiri
, and
J.
Ye
,
Adv. Mater.
24
,
229
251
(
2012
).
2.
S.
Ouyang
and
J.
Ye
,
J. Am. Chem. Soc.
133
,
7757
7763
(
2011
).
3.
Y.
Bi
,
S.
Ouyang
,
N.
Umezawa
,
J.
Cao
, and
J.
Ye
,
J. Am. Chem. Soc.
133
,
6490
6492
(
2011
).
4.
J. S.
Jang
,
H. G.
Kim
, and
J. S.
Lee
,
Catal. Today
185
,
270
277
(
2012
).
5.
J.
Ran
,
J.
Zhang
,
J.
Yu
,
M.
Jaroniec
, and
S. Z.
Qiao
,
Chem. Soc. Rev.
43
,
7787
7812
(
2014
).
6.
S.
Ouyang
,
H.
Tong
,
N.
Umezawa
,
J.
Cao
,
P.
Li
,
Y.
Bi
,
Y.
Zhang
, and
J.
Ye
,
J. Am. Chem. Soc.
134
,
1974
1977
(
2012
).
7.
T.
Simon
,
N.
Bouchonville
,
M. J.
Berr
,
A.
Vaneski
,
A.
Adrović
,
D.
Volbers
,
R.
Wyrwich
,
M.
Döblinger
,
A. S.
Susha
,
A. L.
Rogach
,
F.
Jäckel
,
J. K.
Stolarczyk
, and
J.
Feldmann
,
Nat. Mater.
13
,
1013
1018
(
2014
).
8.
H.
Sheng
,
H.
Ji
,
W.
Ma
,
C.
Chen
, and
J.
Zhao
,
Angew. Chem., Int. Ed.
52
,
9686
9690
(
2013
).
9.
A.
Primo
,
A.
Corma
, and
H.
Garcia
,
Phys. Chem. Chem. Phys.
13
,
886
910
(
2011
).
10.
L.
Liu
,
P.
Li
,
B.
Adisak
,
S.
Ouyang
,
N.
Umezawa
,
J.
Ye
,
R.
Kodiyath
,
T.
Tanabe
,
G. V.
Ramesh
,
S.
Ueda
, and
H.
Abe
,
J. Mater. Chem. A
2
,
9875
9882
(
2014
).
11.
C.
Wang
and
D.
Astruc
,
Chem. Soc. Rev.
43
,
7188
7216
(
2014
).
12.
L.
Liu
,
S.
Ouyang
, and
J.
Ye
,
Angew. Chem., Int. Ed.
52
,
6689
6693
(
2013
).
13.
S.
Linic
,
P.
Christopher
, and
D. B.
Ingram
,
Nat. Mater.
10
,
911
921
(
2011
).
14.
Z.
Lou
,
Z.
Wang
,
B.
Huang
, and
Y.
Dai
,
ChemCatChem
6
,
2456
2476
(
2014
).
15.
Z.
Zheng
,
B.
Huang
,
X.
Qin
,
X.
Zhang
,
Y.
Dai
, and
M.-H.
Whangbo
,
J. Mater. Chem.
21
,
9079
9087
(
2011
).
16.
C.
Clavero
,
Nat. Photonics
8
,
95
103
(
2014
).
17.
H.
Inouye
,
K.
Tanaka
,
I.
Tanahashi
, and
K.
Hirao
,
Phys. Rev. B
57
,
11334
11340
(
1998
).
18.
L.
Du
,
A.
Furube
,
K.
Hara
,
R.
Katoh
, and
M.
Tachiya
,
J. Photochem. Photobiol., C
15
,
21
30
(
2013
).
19.
A.
Furube
,
L.
Du
,
K.
Hara
,
R.
Katoh
, and
M.
Tachiya
,
J. Am. Chem. Soc.
129
,
14852
14853
(
2007
).
20.
J. B.
Priebe
,
M.
Karnahl
,
H.
Junge
,
M.
Beller
,
D.
Hollmann
, and
A.
Bruckner
,
Angew. Chem., Int. Ed.
52
,
11420
11424
(
2013
).
21.
J. Q.
Yan
,
G. J.
Wu
,
N. J.
Guan
, and
L. D.
Li
,
Chem. Commun.
49
,
11767
11769
(
2013
).
22.
Z.
Zheng
,
T.
Tachikawa
, and
T.
Majima
,
J. Am. Chem. Soc.
136
,
6870
6873
(
2014
).
23.
C.
Colbeau-Justin
,
M.
Kunst
, and
D.
Huguenin
,
J. Mater. Sci.
38
,
2429
2437
(
2003
).
24.
Y.
Kwon
,
S. C. S.
Lai
,
P.
Rodriguez
, and
M. T. M.
Koper
,
J. Am. Chem. Soc.
133
,
6914
6917
(
2011
).
25.
S. C. S.
Lai
,
S. E. F.
Kleijn
,
F. T. Z.
Öztürk
,
V. C.
van Rees Vellinga
,
J.
Koning
,
P.
Rodriguez
, and
M. T. M.
Koper
,
Catal. Today
154
,
92
104
(
2010
).
26.
Y.
Kwon
,
K. J. P.
Schouten
, and
M. T. M.
Koper
,
ChemCatChem
3
,
1176
1185
(
2011
).
27.
J.
Hernández
,
J.
Solla-Gullón
,
E.
Herrero
,
A.
Aldaz
, and
J. M.
Feliu
,
Electrochim. Acta
52
,
1662
1669
(
2006
).
28.
B.
Beden
,
I.
Çetin
,
A.
Kahyaoglu
,
D.
Takky
, and
C.
Lamy
,
J. Catal.
104
,
37
46
(
1987
).
29.
A.
Santasalo-Aarnio
,
Y.
Kwon
,
E.
Ahlberg
,
K.
Kontturi
,
T.
Kallio
, and
M. T. M.
Koper
,
Electrochem. Commun.
13
,
466
469
(
2011
).
30.
T.
Sheng
,
W.-F.
Lin
,
C.
Hardacre
, and
P.
Hu
,
J. Phys. Chem. C
118
,
5762
5772
(
2014
).
31.
Y.
Xu
and
M. A. A.
Schoonen
,
Am. Mineral.
85
,
543
556
(
2000
), available at: http://ammin.geoscienceworld.org/content/85/3-4/543.abstract.
32.
R.
Van de krol
,
Y.
Liang
, and
J.
Schoonman
,
J. Mater. Chem.
18
,
2311
2320
(
2008
).
33.
A. V.
Puga
,
A.
Forneli
,
H.
García
, and
A.
Corma
,
Adv. Funct. Mater.
24
,
241
248
(
2014
).
34.
See supplementary material at http://dx.doi.org/10.1063/1.4921783 for detailed experimental methods and data.
35.
V.
Subramanian
,
E. E.
Wolf
, and
P. V.
Kamat
,
J. Am. Chem. Soc.
126
,
4943
4950
(
2004
).
36.
M. M.
Alvarez
,
J. T.
Khoury
,
T. G.
Schaaff
,
M. N.
Shafigullin
,
I.
Vezmar
, and
R. L.
Whetten
,
J. Phys. Chem. B
101
,
3706
3712
(
1997
).
37.
K.
Yamada
,
K.
Miyajima
, and
F.
Mafuné
,
J. Phys. Chem. C
111
,
11246
11251
(
2007
).
38.
Y.
Tian
and
T.
Tatsuma
,
J. Am. Chem. Soc.
127
,
7632
7637
(
2005
).
39.
D. B.
Ingram
and
S.
Linic
,
J. Am. Chem. Soc.
133
,
5202
5205
(
2011
).
40.
T. K.
Townsend
,
N. D.
Browning
, and
F. E.
Osterloh
,
Energy Environ. Sci.
5
,
9543
9550
(
2012
).

Supplementary Material