Fabrication of CdS/TiO2 nano-bulk composite photocatalysts for hydrogen production from aqueous H2S solution under visible light

Article abstract of DOI:10.1016/j.cplett.2006.05.031

CdS/TiO₂ nano-bulk composite (NBC) photocatalyst of a new configuration consisting of bulky CdS with high crystallinity decorated with nanosized TiO₂ particles exhibited an unprecedented high rate of hydrogen production under visible light irradiation (λ ≥ 420 nm) from water containing sulfide and sulfite ions as hole scavengers. The fabrication method of the composite photocatalysts was critical for their performance. It was demonstrated that this active composite photocatalyst could be applied for practical processes of simultaneous hydrogen production and H₂S removal in the treatment of Claus plant tail gas streams or vents of hydrodesulfurization plants.

Full text of DOI:10.1016/j.cplett.2006.05.031

Chemical Physics Letters 425 (2006) 278–282
www.elsevier.com/locate/cplett
Fabrication of CdS/TiO nano-bulk composite photocatalysts
2
for hydrogen production from aqueous H S solution under visible light
2
a
b,1
b,1
a,
*
Jum Suk Jang , Wei Li , Se Hyuk Oh , Jae Sung Lee
a
Department of Chemical Engineering and School of Environmental Science and Engineering, Pohang University of Science and Technology,
San 31 Hyoja-dong, Pohang, Gyungbuk 790-784, Republic of Korea
b
R&D Center, General Motors, Michigan, 480-106-185, 30500 Mound Rd. Warren, MI 48090-9055, USA
Received 21 February 2006; in final form 10 May 2006
Available online 13 May 2006
Abstract
CdS/TiO nano-bulk composite (NBC) photocatalyst of a new configuration consisting of bulky CdS with high crystallinity decorated
2
with nanosized TiO particles exhibited an unprecedented high rate of hydrogen production under visible light irradiation (k P 420 nm)
2
from water containing sulfide and sulfite ions as hole scavengers. The fabrication method of the composite photocatalysts was critical for
their performance. It was demonstrated that this active composite photocatalyst could be applied for practical processes of simultaneous
hydrogen production and H
2
S removal in the treatment of Claus plant tail gas streams or vents of hydrodesulfurization plants.
ꢀ
2006 Elsevier B.V. All rights reserved.
1
. Introduction
composite (NBC) photocatalyst consisting of bulky CdS
with a high crystallinity decorated with nanosized TiO2
particles. This configuration of the composite photocata-
lyst exhibited an unprecedented high rate of hydrogen pro-
duction under visible light (k P 420 nm) from water
containing sulfide and sulfite as sacrificial reagents. As a
practical application of this highly active catalyst, the
simultaneous hydrogen production and decomposition of
H S was performed. Our CdS/TiO NBC photocatalyst
Visible light-driven photocatalysts that could produce
hydrogen from water splitting under solar light have been
actively sought [1–5]. CdS has an ideal band gap energy
(
2.4 eV) and band positions that can drive both oxidation
and reduction of water under visible light irradiation.
Yet, pure CdS is usually not very active in hydrogen pro-
duction, and subject to photo-corrosion [6]. In attempt to
improve the photo-activity and photo-stability, CdS has
been combined with other materials such as ZnO, TiO2
2
2
showed high rates of hydrogen production from H S dis-
2
solved in alkaline water. In both cases of hydrogen produc-
and LaMnO or intercalated into layered compounds [7–
tion from water and decomposition of H S, the catalyst
3
2
1
0]. We have found that the fabrication method of these
was much more active than other configurations of the
composite photocatalysts is critical for their performance.
composite, i.e., nano-CdS/bulk-TiO or nano-CdS/nano-
2
Usually, nanosized CdS and TiO particles are combined
TiO coprecipitates [11,12].
2
2
to obtain the composite in order to maximize the surface
areas. Here, we report a new method of fabricating a highly
2. Experimental
efficient composite photocatalyst, i.e., CdS/TiO nano-bulk
2
To prepare CdS photocatalyst, a stoichiometric amount
of Na S aqueous solution was added drop-by-drop to
*
2
Corresponding author. Fax: +82 562 279 5799/5528.
E-mail addresses: se.h.oh@gm.com (S.H. Oh), jlee@postech.ac.kr (J.S.
Cd(NO ) dissolved in isopropyl alcohol. Precipitated pow-
3
2
der was filtered, dried, and then calcined at 1073 K for 1 h
under He flow to increase the crystallinity of CdS. To
Lee).
1
Fax: +1 586 986 8694.
0
009-2614/$ - see front matter ꢀ 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.cplett.2006.05.031

J.S. Jang et al. / Chemical Physics Letters 425 (2006) 278–282
279
fabricate CdS(bulk)/TiO NBC photocatalyst, the bulky
3. Results and discussion
2
CdS catalyst was stirred in isopropyl alcohol and titanium
isopropoxide (in a mole ratio of Ti to CdS of 4) and H O
Fig. 1 shows X-ray diffraction patterns of the samples.
2
were added drop-by-drop. The prepared composite powder
was filtered, dried, and calcined at 673 K for 2 h under air
The crystallinity of CdS in CdS(bulk)/TiO NBC photocat-
2
alyst was as good as the single phase CdS calcined at
1073 K, considering its concentration (30 wt%) in the com-
posite. The pure CdS showed well-developed hexagonal
flow to increase the crystallinity of TiO . For comparison,
2
the opposite configuration of nano-CdS decorating bulk
TiO was also prepared. Thus, on bulk TiO calcined at
phase. The TiO phase was anatase both in single and
2
2
2
8
73 K, CdS was precipitated from Na S and Cd(NO )
NBC photocatalysts. In the configuration of CdS/TiO (-
2
3 2
2
and calcined at 673 K for 2 h under air flow. The crystalline
phases of the products were determined by X-ray diffrac-
tometer (Mac Science Co., M18XHF). The optical proper-
ties were analyzed by UV–Vis diffuse reflectance
spectrometer (Shimadzu, UV 2401). Morphologies of phot-
ocatalysts were investigated by transmission electron
microscope (JEOL JEM 2010F, Field Emission Electron
Microscope).
The photocatalytic reactions were carried out at room
temperature under normal pressure in a closed circulation
system using a Hg-arc lamp (500 W) equipped with UV
cut off filter (k P 420 nm). Before reaction, 1 wt% of Pt
was deposited on photocatalysts by photodeposition
method under visible light (k P 420 nm). TEM examina-
tion after Pt deposition showed that Pt particles were
bulk), the bulky TiO2 was well-developed anatase, but
CdS was in pure cubic phase. The morphology of CdS/
TiO composite photocatalysts was observed by TEM as
2
shown in Fig. 1. In CdS(bulk)/TiO , CdS particles of ca.
2
1–2 lm was decorated with TiO nanoparticles of ca. 10–
2
20 nm. In CdS/TiO (bulk), the morphology is similar, but
2
now 50–100 nm TiO was decorated with 20 nm CdS.
2
The diffuse reflection spectra for these catalysts are
shown in Fig. 2. The bulky CdS photocatalyst showed a
sharp edge at 570 nm, while that of TiO was observed at
2
390 nm. The spectra of both CdS/TiO composite photo-
2
catalysts showed a combination of these two spectra,
although CdS in composite photocatalysts showed a
blue-shift of the main absorbance edge to 550 nm. This is
in agreement with the previous report by Kisch et al. that
the band gap of CdS employed in composite photocatalyst
is shifted by an electronic semiconductor-support interac-
tion (SEMSI) [13,14].
deposited randomly on both CdS and TiO . The reaction
2
was performed with 0.1 g of photocatalyst in 100 mL aque-
ous solution containing 0.1 M Na S and 0.02 M Na SO .
2
2
3
Hydrogen production from H S solutions was investigated
The rate of hydrogen production was measured from an
2
under the same conditions, but without the sacrificial
reagents. The reaction solutions were prepared by dissolv-
aqueous solution containing 0.1 M Na S and 0.02 M
2
Na SO as sacrificial reagents under visible light irradiation
2
3
ing H S up to saturation in NaOH solution (0, 0.1, and
(k P 420 nm). The time courses of hydrogen production
are shown in Fig. 3. As previously reported [15], there
was an apparent induction period of 1–2 h before the
steady state hydrogen evolution rates of bulky CdS and
2
1
M). The evolved amounts of H were analyzed by gas
2
˚
chromatography (TCD, molecular sieve 5-A column and
Ar carrier).
Fig. 1. X-ray diffraction patterns (XRD) of (a) TiO
2 2 2
calcined at 673 K, (b) CdS(bulk)/TiO , (c) CdS/TiO (bulk), and (d) CdS calcined at 1073 K. TEM
images of (b) and (c) show morphologies of the composite samples.

280
J.S. Jang et al. / Chemical Physics Letters 425 (2006) 278–282
Table 1
2
Rates of H evolution from an aqueous electrolyte solution over CdS and
CdS/TiO composite photocatalysts
2
Photocatalysts
BET area
Bandgap energy
(eV) ab (nm)
H
2
evolution
2
ꢀ1
(
m /g)
(lmol h
)
E
g
k
CdS-673 K
CdS-1073 K
CdS(bulk)/TiO
28.8
<1.0
97.0
26.0
2.18
2.18
2.25
2.25
570
570
550
550
6.9
96.4
422.4
1.3
2
2
CdS/TiO (bulk)
Catalyst; 0.1 g, light source; 500 W Hg-arc lamp (Oriel) with UV cut-off
filter (k P 420 nm), 1 wt% of Pt was deposited on photocatalysts by
photodeposition method under visible light (k P 420 nm) and reaction
was performed in 100 mL electrolyte solution.
CdS appears to be suppressed due to protection by TiO2
nanoparticles decorating the bulk CdS surface.
The average rates of hydrogen evolution during the first
5
h are compared in Table 1 together with other physical
properties. The pure CdS calcined at 1073 K showed much
higher activity than CdS calcined at 673 K, despite of its
much reduced surface area. High temperature treatment
induces the formation of the hexagonal wurtzite phase
(Fig. 1), which is known to be the more active phase than
the low-temperature, cubic zinc blend phase [16]. In addi-
tion, heat treatment could heal surface defects in the crys-
tallites, which are responsible for the energy-wasteful
photoelectron-hole recombination. Now, the CdS(bulk)/
Fig. 2. Diffuse reflection spectra of (a) TiO
2
calcined at 673 K, (b)
CdS(bulk)/TiO , (c) CdS/TiO (bulk), and (d) CdS calcined at 1073 K.
2
2
TiO composite photocatalyst shows even higher photocat-
2
alytic activity (by a factor of ca. 4.4) than that of active sin-
gle phase CdS photocatalyst having similar crystallinity.
The composite catalyst also showed the highest surface
area although its correlation with photocatalytic activity
is not apparent as in the case of two CdS samples.
As inferred in other composite photocatalysts [17,18],
the superior activity of the composite photocatalyst is con-
sidered to be due to faster charge separation compared to
CdS alone upon initial light absorption, caused by fast dif-
fusion of photoelectrons generated in bulky CdS toward
surrounding TiO2 nanoparticles as schematically repre-
sented in Fig. 4. However, the configuration containing
highly-crystalline, hexagonal-phase CdS is essential for
high photocatalytic activity since the reverse configuration
Fig. 3. Photocatalytic hydrogen production of (a) CdS calcined at 673 K,
b) CdS(bulk) calcined at 1073 K, (c) CdS/TiO (bulk), (d) CdS(bulk)/
TiO . Catalysis: 0.1 g loaded 1 wt% Pt, Electrolyte: 0.1 M Na S + 0.02 M
Na SO
of CdS/TiO (bulk) containing nanosized CdS particles of
2
(
2
the cubic phase are far less active (by a factor of 325!). It
2
2
is also much more active than nano-CdS/nano-TiO copre-
2
3
.
2
cipitates reported in the literature [19,20]. It is remarkable
that the fabrication method of the composite has such a
dramatic effect despite the composition was almost the
same. To the best of our knowledge, the rate of hydrogen
evolution over CdS(bulk)/TiO2 composite photocatalyst
represents one of the highest values ever reported under
similar reaction conditions for any visible light active phot-
ocatalysts. The activity for CdS(bulk)/TiO2 composite
photocatalyst for hydrogen production is at least an-order-
of-magnitude higher than other CdS-containing composite
systems made mostly of CdS nanoparticles [21–25]. Only
complicated solid solution of ZnS–CuInS –AgInS showed
CdS/TiO (bulk) photocatalysts were established. Yet,
2
CdS(bulk)/TiO photocatalyst did not show such an induc-
2
tion period, and showed the highest rate of hydrogen
production.
The surface sulfur species was analyzed by XPS before
photoreaction (not shown here). The S2p peaks indicated
2
4
ꢀ
that CdS calcined at 1073 K contained SO as an impu-
2
ꢀ
rity, whereas CdS(bulk)/TiO contained only sulfide (S
)
2
species. This surface oxidation may be responsible for the
induction period for the single-component CdS photocata-
2
2
lysts. In CdS(bulk)/TiO catalyst, the surface oxidation of
the higher activity [26].
2

J.S. Jang et al. / Chemical Physics Letters 425 (2006) 278–282
281
2 2
Fig. 4. A new configuration model consisting of bulky CdS with high crystallinity decorated with nanosized TiO particles. The possible role of TiO
nanoparticles is to provide sites for collecting the photoelectrons generated from CdS, enabling thereby an efficient electron–hole separation as depicted.
Hydrogen production from water in the presence of sac-
rificial reagents is the standard method of evaluating the
activity of metal sulfide photocatalysts, yet the method
itself has a limited practical value because the sacrificial
reagents are irreversibly consumed. However, the high
hydrogen production activity over our CdS(bulk)/TiO2
composite photocatalyst could be exploited for the practi-
cal process of simultaneous hydrogen production and
H S removal, e.g., the treatment of Claus plant tail-gas
2
stream or vents of hydrodesulfurization plants [27–30]. In
H S scrubbers used in such industry, alkaline water is com-
2
monly employed for effective H S scrubbing by an acid–
2
base reaction (1) showed below. Note that the soluble
alkali metal sulfides are common sacrificial reagents for
hydrogen production with metal sulfide photocatalysts
and are also employed to improve the activity and stability
Fig. 5. Photocatalytic hydrogen production over various electrolyte
of the sulfide photocatalysts. They are formed in situ here
2 2 3 2 2
S + 0.02 M Na SO , (b) H S dissolved in H O,
solutions; (a) 0.1 M Na
(c) H
Catalyst: CdS(bulk)/TiO
when H S is scrubbed by alkaline water.
2
2
S dissolved in 0.1 M NaOH, and (d) H
2
S dissolved in 1 M NaOH.
The following reactions are known to occur during pho-
2
composite (0.1 g loaded 1 wt% Pt), light source:
Hg-arc lamp (500 W) equipped with UV cut-off filter (k P 420 nm).
tocatalytic H S cleavage:
2
ꢀ
2ꢀ
H S þ 2OH ! S þ 2H O
ð1Þ
ð2Þ
ð3Þ
2
2
ꢀ
þ 2OHꢀ
2
e þ 2H
2
O ! H
! S þ 2e
2
2ꢀ
0
ꢀ
water splitting experiments over sulfide photocatalysts
with Na S/Na SO as sacrificial reagents. As mentioned,
S
2
2
3
0
The net reaction is H S cleavage to H and S , but water
sulfide ion is formed stoichiometrically in situ by reaction
between H S and NaOH. In contrast, H S dissolved in
pure water showed an extremely low photoactivity.
2
2
also contributes to the hydrogen production through the
reaction (2).
2
2
Thus, we dissolved H S in pure water and NaOH
When H S is bubbled into pure water, simple dissolution
2
2
solution. Shown in Fig. 5 are the results of photocata-
lytic hydrogen production from various electrolyte solu-
tions over CdS(bulk)/TiO2 photocatalyst under visible
light irradiation (k P 420 nm for all catalysts). High
rates of hydrogen evolution were observed from H2S
electrolyte dissolved in NaOH. It should be noted that
of H S takes place instead of the chemical reaction
2
observed when an alkali is present. The solubility in
water is as high as ca. 0.1 M (at 298 K), roughly equal
to titrated alkaline solution. But the dissociation of
H S is severely limited as indicated by the pK value of
2
ꢀ
+
2ꢀ
11.96 at 291 K for the reaction of HS « H + S
Thus, the low photocatalytic activity of H S dissolved
.
the initial rates of hydrogen production from H S/
2
2
0
.1 M NaOH/H O and H O/0.1 M Na S/0.02 M Na SO
3
in pure water may originate from the limited availability
of sulfide ions.
2
2
2
2
are very similar. The latter is the typical condition of

2
82
J.S. Jang et al. / Chemical Physics Letters 425 (2006) 278–282
[7] P.V. Kamat, Chem. Rev. 93 (1993) 267.
4
. Conclusions
[8] N. Serpone, E. Borgarello, M. Gr a¨ tzel, J. Chem. Soc. Chem.
Commun. (1983) 342.
9] L. Spanhel, H. Weller, A. Henglein, J. Am. Chem. Soc. 109 (1987)
6632.
In conclusion, we report that the fabrication of
[
CdS(bulk)/TiO NBC photocatalyst based on crystalline
2
CdS decorated with TiO nano-particles is a successful
[10] W. Shangguan, A. Yoshida, J. Phys. Chem. B 106 (2002) 12227.
[11] S.A. Naman, M. Gr a¨ tzel, J. Photochem. Photobiol. A 77 (1994) 249.
2
strategy to develop a highly active photocatalyst for hydro-
gen production from water under visible light irradiation.
This NBC structure results in an efficient charge separa-
tion, caused by fast diffusion of photoelectrons generated
[12] M.M. Taquikhan, R.C. Bhardwaj, C. Bhardwaj, Int. J. Hydrogen
Energ. 13 (1988) 7.
[13] H. Weib, A. Fernandez, H. Kisch, Angew. Chem. Int. Ed. 40 (2001)
3
825.
from bulky CdS toward surrounding TiO nanoparticles,
[14] H. Kisch, H. Weib, Adv. Funct. Mater. 12 (2002) 483.
2
leading to high photocatalytic activity of hydrogen produc-
tion. This active composite photocatalyst could be applied
for practical processes of simultaneous hydrogen produc-
[15] N. Serpone, E. Borgarello, Inorg. Chim. Acta 90 (1984) 191.
[16] M. Matsumura, S. Furukawa, Y. Shao, H. Tsuboumura, J. Phys.
Chem. 89 (1985) 1327.
[17] L. Spanhel, H. Weller, A. Henglein, J. Am. Chem. Soc. 109 (1987)
6632.
tion and H S removal in the treatment of Claus plant tail
2
gas streams or vents of hydrodesulfurization plants.
[18] J.Z. Zhang, J. Phys. Chem. B 104 (2000) 7239.
[
[
[
19] H. Fujii, M. Ohtaki, K. Eguchi, H. Arai, J. Mol. Catal. A 129 (1998)
1.
20] S.V. Tambwekar, D. Venugopal, M. Subrahmanyam, Int. J. Hydro-
gen Energ. 22 (1999) 957.
21] I. Tsuji, H. Kato, H. Kobayashi, A. Kudo, J. Am. Chem. Soc. 120
(2004) 13406.
6
Acknowledgements
This work was supported by General Motors R&D Cen-
ter, the Hydrogen Energy R&D Center (one of the 21st
Century Frontier R&D Program), the Brain Korea 21 Pro-
ject, National R&D Project for Nano Science and Technol-
ogy and National Research Laboratory.
[22] Z. Lei, W. You, M. Liu, G. Zhou, T. Takata, M. Hara, K. Domen, C.
Li, Chem. Commun. (2003) 2142.
[
[
[
23] G. Guan, A. Kida, K. Kusakabe, K. Kimura, X. Fang, T. Ma, E.
Abe, A. Yoshida, Chem. Phys. Lett. 385 (2004) 319.
24] A. Kida, G. Guan, A. Yoshida, Chem. Phys. Lett. 371 (2003)
References
563.
25] W.W. So, K.J. Kim, S.J. Moon, Int. J. Hydrogen Energ. 29 (2004)
229.
[
[
1] A. Kudo, Catal. Surv. Asia 7 (2003) 31.
2] K. Domen, M. Hara, J.N. Kondo, T. Takata, A. Kudo, H.
Kobayashi, Y. Inoue, Korean J. Chem. Eng. 18 (2001) 862.
3] H.G. Kim, D.W. Hwang, J.S. Lee, J. Am. Chem. Soc. 126 (2004)
[26] I. Tsuji, H. Kato, A. Kudo, Angew. Chem. Int. Ed. 44 (2005) 2.
[27] M. Barbeni, E. Pelizzetti, E. Borgarello, N. Serpone, M. Gr a¨ tzel, L.
Balducci, M. Visca, Int. J. Hydrogen Energ. 10 (1985) 249.
[28] C.A. Linkous, N.Z. Muradov, S.N. Ramser, Int. J. Hydrogen Energ.
20 (1995) 701.
[
8912.
[
[
4] H. Kato, K. Asakura, A. Kudo, J. Am. Chem. Soc. 125 (2003) 3082.
5] H.G. Kim, P.H. Borse, W. Choi, J.S. Lee, Angew. Chem. Int. Ed. 44
[29] L.R. Grzyll, J.J. Thomas, R.G. Barile, Int. J. Hydrogen Energ. 14
(1989) 647.
(2005) 4585.
[
6] A.J. Frank, K. Honda, J. Phys. Chem. 86 (1982) 1933.
[30] N. Buhler, K. Meier, J.F. Reber, J. Phys. Chem. 88 (1984) 3261.