You are on page 1of 11

This article appeared in a journal published by Elsevier.

The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elseviers archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright

Author's personal copy

Energy 34 (2009) 16521661

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Development of copper-doped TiO2 photocatalyst for hydrogen production


under visible light
L.S. Yoong a, F.K. Chong a, Binay K. Dutta b, *
a
b

Chemical Engineering Department, University of Technology PETRONAS, Bandar Seri Iskandar, 31750 Tronoh, Perak, Malaysia
Chemical Engineering Department, The Petroleum Institute, P.O. Box 2533, Abu Dhabi, United Arab Emirates

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 16 March 2009
Received in revised form
8 July 2009
Accepted 22 July 2009
Available online 15 August 2009

The advantage of copper doping onto TiO2 semiconductor photocatalyst for enhanced hydrogen generation under irradiation at the visible range of the electromagnetic spectrum has been investigated. Two
methods of preparation for the copper-doped catalyst were selected complex precipitation and wet
impregnation methods using copper nitrate trihydrate as the starting material. The dopant loading
varied from 2 to 15%. Characterization of the photocatalysts was done by thermogravimetric analysis
(TGA), temperature programmed reduction (TPR), diffuse reectance UV-Vis (DR-UV-Vis), scanning
electron microscopy (SEM), Fourier transform infrared (FTIR) spectroscopy and X-ray diffraction (XRD).
Photocatalytic activity towards hydrogen generation from water was investigated using a multiport
photocatalytic reactor under visible light illumination with methanol added as a hole scavenger. Three
calcination temperatures were selected 300, 400 and 500  C. It was found that 10 wt.% Cu/TiO2 calcined
at 300  C for 30 min yielded the maximum quantity of hydrogen. The reduction of band gap as a result of
doping was estimated and the inuence of the process parameters on catalytic activity is explained.
2009 Elsevier Ltd. All rights reserved.

Keywords:
Copper-doped TiO2
Hydrogen production
Solar hydrogen
Photocatalyst

1. Introduction
The current global economy is largely dependent on fossil fuels
which are virtually integral to industry, transport and everyday life.
Because of the depleting reserves and the gross environmental
impact of burning fossil fuels, identication of suitable alternative
and renewable energy sources such as hydrogen is gaining importance. Hydrogen gas is foreseen as the future energy carrier since it
is renewable, it does not generate the greenhouse gas CO2 in
combustion, has a large specic energy density and is easily
convertible to electricity by fuel cells [1]. Although hydrogen is now
used primarily as a bulk reducing agent in petrochemical, food,
electronics and metallurgical processing industries, it is rapidly
emerging as a sustainable and the cleanest energy source for
transportation and utilities industries in general [2,3]. Hydrogen is
produced from fossil fuels as well as from biomass by established
technologies [4] but the conventional sources should be eventually
substituted by alternative sources such as solar energy. In fact, one
of the most promising renewable technologies for the production of
H2 is photochemical and photocatalytic water splitting using solar
energy [2]. Further technological developments in solar hydrogen

* Corresponding author. Tel.: 971 2 6075246; fax: 971 2 6075200.


E-mail address: bdutta@pi.ac.ae (B.K. Dutta).
0360-5442/$ see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.energy.2009.07.024

production will provide a sustainable solution to both energy and


environmental problems [1]. In a recent paper, Penner [5] has
discussed the current technologies, including the solar photocatalytic one, of hydrogen production and their economic aspects.
Guo et al. [6] reviewed some of the recent developments in solar
hydrogen production.
Titanium dioxide or titania (TiO2) is considered as a good
photocatalyst for hydrogen generation because of its excellent
resistance to chemical and photochemical corrosion in aggressive
aqueous environments and due to its activity towards both light
and water. It is also less expensive than many other photo-sensitive
materials [7]. There are three naturally occurring TiO2 phases:
anatase, rutile and brookite. Anatase is the most active phase and is
widely used as a photocatalyst [8]. It is a metastable phase (that is
why it is the most active of the three varieties) of TiO2 which can be
transformed rapidly into the inactive rutile phase at temperatures
above 600  C. However, pure TiO2 is not photoactive under
illumination by visible light (wavelength from 400 to 800 nm). It
can only absorb near-UV range of approximately 388 nm or less.
The band gap for the anatase form is Eg 3.2 eV, and this why it is
excitable only in the UV range. This part of the spectrum accounts
for only 34% of solar radiation [9] and use of UV radiation alone for
photocatalytic splitting of water to hydrogen is not an economically
feasible option. Therefore, in order for TiO2 to harvest the energy
from the visible region of sunlight, the absorption edge of TiO2

Author's personal copy

L.S. Yoong et al. / Energy 34 (2009) 16521661

must be shifted to this region. It is well known that when a semiconductor photocatalyst such as TiO2 is exposed to radiant energy
of suitable wavelength, an electron in the valence band jumps to
the conduction band leaving a hole behind. However, once in the
excited state, the conduction band (CB) electron and the valence
band (VB) hole can recombine very quickly releasing energy in the
form of unproductive heat or photons without undergoing any
photoreaction. This phenomenon explains the slow hydrogen
production by photocatalytic water splitting using TiO2 alone [10].
In order to overcome this problem, numerous studies have been
conducted to enhance the photocatalytic activity of TiO2 by incorporating an adequate amount of a transition metal oxide such as
that of Fe, Zn, Cu, Ni and V [5,1114]. Doping of these metals
essentially lessens the band gap of TiO2 for the photo-excitation
and simultaneously reduces the recombination rate of photogenerated electronhole pairs. Besides doping of the photocatalyst,
addition of a hole scavenger or sacricial electron donor can
enhance the efciency of photoreduction of water to H2. The hole
scavenger reacts irreversibly with the photogenerated holes
suppressing the recombination of electrons and holes on the
semiconductor surface. This leads to a higher rate of photocatalytic
reduction of water to hydrogen [15]. Methanol, ethanol, lactic acid,
formaldehyde, cyanides and ethylenediaminetetraacetic acid
(EDTA) have been tested and found to be effective for the
enhancement of hydrogen production and other photocatalytic
reduction reactions [16].
Copper doping has been less explored compared to other transition metals mentioned above. Also it appears from the limited
literature reports available on copper doping that the doped catalysts have been tested only in the UV or near-UV range of radiation
[13,17]. In the present study we have prepared copper-doped TiO2
photocatalysts using two different methods and tested their efciency for hydrogen generation. Thus, the principal objective of the
work is to develop Cu/TiO2 photocatalysts for hydrogen production
from water under irradiation with visible light. Since we intended
to shift the absorption edge of TiO2 to the visible region, the
photoreaction was conducted using a 500-W halogen lamp to
simulate visible light. Characterization of the catalysts was done
following standard procedures. Another set of photocatalyst was
prepared by nickel-doping of TiO2 following the same procedure in
order to test its effectiveness to reduce the band gap.
2. Experimental
2.1. Catalyst preparation
The effectiveness of a doped photocatalyst depends upon the
selection and loading of the dopant as well as the preparation
method. The physical form of the dopant and its degree of dispersion on the base material (here TiO2) are other important factors in
this respect. Accordingly, we have prepared a set of photocatalysts
with the dopant metal loading of 2, 5, 10 and 15 wt.% on Degussa
P25 TiO2 photocatalyst (Aerosil, Germany) containing predominantly anatase phase and having a specic surface area of 50 m2/g.
The catalyst samples were prepared using two methods complex
precipitation and wet impregnation. Copper nitrate trihydrate,
Cu(NO3)2.3H2O (Acros, >98% purity), was used as the dopant metal
precursor for both the preparation methods. Glycerol (Systerm, 95%
purity) was used as the complexing agent for copper. Sodium
hydroxide, NaOH (Merck, 95%), was used as the precipitating agent
in the complex precipitation method.
2.1.1. Complex precipitation method
Glycerol was added to an aqueous solution of Cu(NO3)2.3H2O
(Cu:glycerol mole ratio 1:2) to generate a copperglycerol

1653

complex. This was followed by addition of TiO2 to this solution with


continuous stirring to form a suspension. The copperglycerol
complex was precipitated on TiO2 particles by adding NaOH
(0.25 M) dropwise into the suspension with constant stirring. The
precipitate formed was further stirred intensely for 30 min prior to
ltering and then drying overnight at 70  C. Thermal decomposition on the raw catalyst was conducted using a thermal gravimetric
analyzer prior to calcination in a mufe furnace at 300  C, 400  C or
500  C for 30 min. A catalyst sample is designated as mCuGTa_b,
where m is the percentage of Cu loading; CuGT refers to copper
glycerol complex precipitation method, a is the calcination
temperature (100  C) and b is the duration of calcination (in
min)1. The accuracy of measurement of weight was up to 0.01 mg
and that of temperature was 1  C.
2.1.2. Wet impregnation method
Catalysts were also prepared by aqueous wet impregnation of
TiO2 using the same copper salt as the precursor for copper loading.
Wet impregnation is a widely used catalyst preparation technique
where the precursor material is dissolved in a solvent and mixed
with the support. The solvent is then removed by evaporation when
the precursor gets deposited on the base material, i.e., the support.
Accordingly, a solution of copper nitrate in distilled water was
stirred for 1 h with TiO2 added to it. The solvent was evaporated
slowly in a water bath at 80  C. As in the complex precipitation
method, 215wt% of copper was introduced as the dopant. The
material was oven-dried at 120  C and then calcined at 300 , 400
or 500  C for 30 min. Catalysts prepared using this method are
designated as mCuTa_b, where m is the percentage of Cu loading (%)
[Cu copper; T titania]; CuT refers to the wet impregnation
method, a is the calcination temperature (100  C) and b is the
duration of calcination as before (in min)2. The catalysts prepared
are listed in Table 1.

2.2. Catalyst characterization


The catalyst samples prepared by the above two methods were
activated by the calcination process. In order to determine the
appropriate temperature and duration of calcination, preliminary
experiments using a thermogravimetric analyzer was conducted.
The thermogravimetric analysis (TGA) was carried out using the
PerkinElmer TG system to determine the thermal stability of the
catalyst materials prior to calcination. The rate of heating was
maintained at 20  C/min, and the measurement was carried out
from 30  C to 800  C with air owing at a rate of 20 mL/min. TGA
results are reported as thermograms which are plots of the
percentage decomposition of the catalyst versus temperature.
Based on the thermograms, the calcination temperatures for activation of the catalysts prepared by both the methods (precipitation
and wet impregnation) were selected as 300  C, 400  C and 500  C
for a duration of 30 min.
It is important to characterize the calcined catalysts in order to
determine mainly their chemical and physical properties and then
to relate these properties to their photocatalytic performance.
Catalyst characterization acts as a ngerprint to record the
uniqueness of the material, the functional groups and the molecular properties so that two catalyst materials belonging to the same
class can be compared for their performance with respect to the

1
For example, 10CuGT3_30 means a catalyst with 10% copper loading prepared
by the complex precipitation with glycerol, calcined at 300  C (i.e., 3  100  C) for
30 min.
2
Similarly, 10CuT4_30 means a catalyst with copper loading prepared by wet
impregnation, calcined at 400  C (4  100  C) for 30 min.

Author's personal copy

1654

L.S. Yoong et al. / Energy 34 (2009) 16521661

Table 1
Details of the catalysts prepared.
Metal loading, wt.%

2
5
10
15

Precipitation method

Wet impregnation method

Temp.: 300  C

400  C

500  C

300  C

400  C

500  C

2CuGT3_30
5CuGT3_30
10CuGT3_30
15CuGT3_30

2CuGT4_30
5CuGT4_30
10CuGT4_30
15CuGT4_30

2CuGT5_30
5CuGT5_30
10CuGT5_30
15CuGT5_30

2CuT3_30
5CuT3_30
10CuT3_30
15CuT3_30

2CuT4_30
5CuT4_30
10CuT4_30
15CuT4_30

2CuT5_30
5CuT5_30
10CuT5_30
15CuT5_30

results of characterization. The characterization methods used in


this study include temperature programmed reduction (TPR),
Fourier transform infrared spectroscopy (FTIR), diffuse reectance
spectroscopy (DR-UV-Vis), X-ray diffraction (XRD) and scanning
electron microscopy (SEM).
TPR allows determination of the type and amount of different
reducible species present in a catalyst sample and also helps in
identication of characteristic reduction peaks of these species [18].
We conducted the TPR analysis using the Thermo Finnigan TPDRO
1000 instrument. A catalyst sample was loaded into a quartz U-tube
reactor that is a part of this instrument. Prior to reduction, the
sample (0.1 g0.2 g) was pretreated in a nitrogen stream at a ow
rate of 20 mL/min (measurement accuracy 2%). The temperature was increased from 40  C to 110  C at a rate of 10  C/min and
maintaining at the desired level for 10 min. This was followed by
reduction in a stream of 5% H2 and 95% N2 from 40  C to 400  C with
a temperature ramp rate of 10  C/min. The sample was held at the
nal temperature for 30 min. Results from TPR were recorded as
reduction proles where the hydrogen consumption was monitored with a thermal conductivity detector (TCD). The total amount
of hydrogen consumed (mmol/g catalyst) could be calculated by
integrating the areas under all the reduction peaks present.
Infrared spectra of the catalyst samples were obtained using
PerkinElmer Spectrum One spectrometer. A sample for FTIR analysis was prepared by grinding and mixing 1 mg of the catalyst with
200 mg of IR-grade KBr and then the sample was pressed into
a pellet using a hydraulic press. The FTIR spectrum of the pellet,
taken over a wavenumber range of 450 cm14000 cm1, was
recorded as the percentage of transmittance (%T) versus wavenumber. FTIR spectra are useful for the determination of the
functional groups (such as CH3, NO
3 , OH, CO, etc.) present in the
catalysts before and after the calcination process. The functional
groups are identied by characteristic peaks in the spectrum.
Diffuse reectance UV-Vis spectra were recorded on a Perkin
Elmer Lambda 900 instrument with an integrating sphere attachment using BaSO4 powder as an internal reference. The layer of
powder sample was made sufciently thick such that all incident
light was absorbed or scattered before reaching the back surface of
the sample holder. Typically a thickness of 13 mm is required. The
diffuse reectance spectra were plotted as the KubelkaMunk
function or remission, F(R), versus wavelength. Based on the
KubelkaMunk equation, F(R) (1  R)2/2R, where reectance,
R Rsample/Rreference. The relative position of the catalysts absorption edge compared to TiO2 in the diffuse reectance spectra is an
important indication whether the absorption edge has been
successfully shifted to the visible region. Since the major objective
of this work is to modify anatase TiO2 by doping with appropriate
metals so as to enhance absorption of visible light, DR-UV-Vis
spectra of the catalyst material are very important for its characterization. The band gaps (Eg) for all the catalysts were determined
from the extrapolation of the linear t for the Tauc plot onto the
photon energy axis (please refer to Section 3.3).
The morphologies of the catalysts were determined using SEM
(LEO 1543). The samples were coated with a layer of platinum
palladium prior to scanning at 100 K magnication. An SEM picture

is also expected to show how well the dopant material is dispersed


on the base material, viz TiO2, in a catalyst sample. Powder-XRD
(Bruker D8 Advance) was conducted on the catalysts with CuKa
radiation of 40 kV, 40 mA; 2q angles from 10 to 80 and scan speed
of 4 /min in order to identify the type of Cu species and also the
TiO2 phases present. The XRD peaks are compared with standards
in order to determine the species present in a sample.
2.3. Photocatalytic activity
The photocatalytic reaction study was performed at room
temperature (23  C) using a multiport photocatalytic reactor
schematically shown in Fig. 1. It was made of a thick acrylic block (1)
in which four reactor ports (40 mL volume) were cut out as shown
in the gure. Each reactor port was covered with a quartz disk (2)
held in place by a screwed square metal plate (3) open at the center.
The acrylic block was xed to a base plate by screws (4). The
photocatalytic reactor was tted with pressure gauges (5) and oneway valves (6). A 500-W halogen lamp with a reecting shade was
used as the light source simulating solar radiation. It was positioned above the multiport reactor at a height of 15 cm. The total
radiation ux was 368 W/m2 (the accuracy of measurement of
radiation ux was 5%). The catalyst was dispersed in a mixture of
8 mL of distilled water and 0.5 mL of methanol that acts as a hole
scavenger as explained before. The gas evolved was collected in
a vertical graduated glass tube over a period of 1 h by water
displacement. The gas volume was measured with an accuracy of
2%. The gaseous products were analyzed using a mass spectrometer with N2 as the carrier owing at 20 mL/min.

Fig. 1. Schematic of the multiport photocatalytic reactor.

Author's personal copy

L.S. Yoong et al. / Energy 34 (2009) 16521661

1655

100

Relative weight, wt %

Relative weight, wt %

100
98
96
94

92
30

98
96
94
92
90

88
130

230

330

430

530

630

30

730

130

230

Tempreature, C

330

430

530

630

730

Temperature, C

Fig. 2. The weight loss prole of the fresh copper supported TiO2 catalyst prepared using (a) precipitation method and (b) wet impregnation method.

3. Results and discussion


3.1. TGA and TPR
The TGA of copper-doped on TiO2 catalyst samples showed the
weight loss pattern of the material prepared freshly by the complex
precipitation (Fig. 2a) and by the wet impregnation methods
(Fig. 2b). The thermogram for 10CuT showed two decomposition
steps one from 30  C to 110  C and the other from 200  C to
400  C. These are attributed to moisture loss and decomposition of
copper nitrate salt, Cu(NO3)2 (Equation 1), respectively. The total
weight loss was about 11.5%. As for the thermogram for 10CuGT, the
rst weight loss step, which depicts the evaporation of the physically retained or adsorbed water occurring between 30  C and
150  C, was slower than that for 10CuT. The rapid weight loss (ca.
8%) displayed by 10CuGT between 150  C and 330  C represent the
decomposition of the copper hydroxide, Cu(OH)2, to form copper
oxide, CuO (Equation 2). The TG curves were used to obtain information about the optimum calcination temperature for the
samples.

CuNO3 2 / CuOs 2NO2 g O2 g

(1)

CuOH2 / CuOs H2 O

(2)

The TPR proles of the catalysts prepared by complex precipitation


(Fig. 3a) displayed dominant reduction peaks that appear at 251  C,
264  C and 285  C for bulk CuO from 10CuGT calcined at 300 , 400
and 500  C, respectively, with successively decreasing hydrogen
consumption as shown in Table 2. The shoulders at lower reduction
temperature (ca. 220  C) are related to reduction of small CuO
species with lesser interaction with the support [19].

3.2. FTIR
Fig. 4 shows the FTIR transmission spectra of the 10CuGT
(Fig. 4a) and 10CuT (Fig. 4b) catalysts calcined at 300  C, 400  C and
500  C for 30 min. In all the spectra, the absorption peaks around
1600 cm1 and 3400 cm1 are attributed to the OH bending and
stretching, respectively [21]. The IR band observed from 400 to
900 cm1 corresponds to the TiO stretching vibrations [22]. On
the other hand, the absorption band at 1384 cm1 which is
attributed to the presence of nitrate (NO
3 ) group could not be
detected on the spectra indicating that the calcination process was
able to completely remove the NO
3 group from the raw catalysts.
This is in agreement to the results obtained from TPR analysis (as

40000
10CuGT3_30

30000

Hydrogen consumed, mol/g

Hydrogen consumed, mol/g

As for the catalyst prepared by wet impregnation, the TPR


proles (Fig. 3b) showed two reduction peaks for 10CuT calcined at
400  C (reduction peaks at 172  C and 241  C) and at 500  C
(reduction peaks at 169  C and 241  C). On the other hand, only one
dominant reduction peak at 256  C was observed for Cu/TiO2
calcined at 300  C (there is a shoulder ca. 170  C). The lower
reduction temperature can be ascribed to reduction of small CuO
particles having little or no interaction with the support [19] while
higher reduction temperature observed in all samples prepared by
both methods is due to the large particle formation of bulk CuO
having stronger interaction with the support [20]. This is indicated
by the amount of H2 consumed which decreases as the calcination
temperature increases (Table 2). It can be concluded from Table 2
that a lower calcination temperature for both the preparation
methods results in better dispersion of copper species on the
support thus leading to higher H2 consumption for reduction of
the CuO.

10CuGT4_ 30
10CuGT5_ 30

20000

10000

0
0

100

200
300
Temperature, C

400

30000
10CuT3_30
10CuT4_ 30

20000

10CuT5_ 30
10000

0
0

200

400

600

800

Temperature, C

Fig. 3. The TPR proles of H2 consumed for catalysts prepared by (a) precipitation method and (b) wet impregnation method (calcination temperature: 300  C, 400  C and 500  C).

Author's personal copy

1656

L.S. Yoong et al. / Energy 34 (2009) 16521661

Table 2
Hydrogen consumption in TPR for the catalysts (precipitation and wet impregnation
methods).
Precipitation method

1
14
13

Wet impregnation method

12

Catalyst

10CuGT3_30
10CuGT4_30
10CuGT5_30

Reduction
peak, ( C)
251
264
285

Amount of H2
adsorbed
(mmol/g)
1429
1310
1210

Catalyst

Reduction
peak, ( C)

10CuT3_30
10CuT4_30
10CuT5_30

256
172, 241
169, 241

Amount of H2
adsorbed
(mmol/g)

11

10CuGT4_30

10

10CuGT3_30

2940
1166
1204

F(R)

7
6

10CuGT5_30

well as from XRD described later) where copper exists as CuO after
the calcination process.

4
3
2

3.3. DR-UV-Vis

TiO 2

The DR-UV-Vis spectra of the catalyst samples as well as of pure


TiO2 are depicted in Fig. 5a and b. The spectra of TiO2 showed an
absorption peak at 388 nm [23] in the UV region. When doped with
10 wt.% copper onto TiO2, considerable shift of the peak towards the
visible range at around 400800 nm occurred for all the samples.
According to Colon et al. [24], the band at 210270 nm would
indicate the O2 (2p)/Cu2 (3d) ligand to metal charge transfer
transition, where the Cu ions occupy isolated sites over the support.
A broad band between 400 and 600 nm is attributed to the presence of Cu1 clusters in partially reduced CuO matrix as well as
(CuOCu)2 clusters. The absorption band at 600800 nm indicates the crystalline and bulk CuO in octahedral symmetry [24]. The
incorporation of copper ion shifts the absorption band to the visible
or even near-infrared range and this promotes the photocatalytic
activity of 10CuGT3_30 and 10CuT3_30. Thus, it can be inferred that
doping with a transition metal such as copper was effective for
visible light response [25] and can play a signicant role in
enhancing the hydrogen evolution [26] using solar energy
economically. The UV-Vis absorption edge and band gap energies of
the samples have been determined from the reectance [F(R)]
spectra using the KM (KubelkaMunk) formalism and the Tauc plot.
For a semiconductor material, a plot of [F(R).hn]n against hn should
show a linear region just above the optical absorption edge for
n if the band gap is a direct transition, or for n 2 if it is indirect
[2729]. Over the linear region of the plots, the relationship can be
described as



FR:hn1=2 K hn  Eg

(3)

where hn photon energy, Eg the band gap energy, and K a


constant characteristic of the semiconductor material.
From Eq. (3) it appears that extrapolation of a Tauc plot to the hn
axis should yield the semiconductor band gap energy. The

0.
190.

25

30

35

40

45

50

55

60

65

70

75

800.

650

700

750

800.0

nm

15
14
13

10CuT4_30

12
11
10
9

10CuT3_30

F(R)

10CuT5_30

6
5
4
3
2

TiO2

1
0.0
190.0

250

300

350

400

450

500

550

600

nm
Fig. 5. (a) The DR-UV-Vis spectra of TiO2 and 10CuGT (complex precipitation). (b) The
Dr-UV-Vis spectra of TiO2 and 10CuT (wet impregnation).

extrapolation lines shown in Fig. 6a and b have been used to


determine the band gaps for the different catalyst samples tested.
The calculated band gap energy for pure TiO2 is also found to be
3.2 eV from the extrapolation of the corresponding plot. The
calculated values of the band gap energy are given in Table 3. All the
catalysts displayed reduction in their band gaps compare to TiO2.
The largest reduction in band gap is observed for 10CuT3_30. This is
in conformity with higher hydrogen generation using the catalyst

10CuGT3_30

10CuT3_30

10CuGT4_30
%T

%T

10CuT4_30

10CuGT5_30

4000.0

3000

10CuT5_30

2000

1500
-1

cm

1000

400.0

4000.0

3000

2000

1500

cm

1000

400.0

-1

Fig. 4. The FTIR spectra of 10 wt.% copper supported catalyst prepared using (a) precipitation method (b) wet impregnation method.

Author's personal copy

L.S. Yoong et al. / Energy 34 (2009) 16521661

8
7
6

[F(R).hv]1/2

1657

(a) 10CuGT4_30

(a)

(b) 10CuGT3_30

(b)

(c) 10CuGT5_30

(c)

(d) TiO2

(d)

3
2
1
0
1.5

2.5

3.5

hv, eV

8
(a)

(a) 10CuT4_30

[F(R).hv]1/2

(b) 10CuT5_30

(c) 10CuT3_30

(d) TiO2

(b)
(c)

4
(d)

3
2
1
0
1.5

2.5

3.5

hv, eV
Fig. 6. a) Plots of transformed KubelkaMunk functions [F(R).hv]1/2 versus hv for
10CuGT (precipitation) and undoped titania samples to estimate band gap energies by
linear extrapolation. (b) Plots of transformed KubelkaMunk functions [F(R).hv]1/2
versus hv for 10CuT (wet impregnation) and undoped titania samples to estimate band
gap energies by linear extrapolation.

containing 10% copper and calcined at 300  C. However, in order to


further enhance the photocatalytic activity, good interaction of CuO
with TiO2 is required which could be conrmed by the SEM
pictures.
3.4. XRD
The XRD patterns of the Degussa P25 and CuO/TiO2 prepared by
using the complex precipitation (10CuGT3_30) as well as the wet
impregnation (10CuT3_30) methods are shown in Fig. 7. The peaks
at 2q 25.34 and 2q 27.42 appear in all samples. These correspond to the main peak of anatase and rutile, respectively [30]. CuO
(tenorite) diffraction peaks appeared near 2q 35.6 and 38.73 for
all the samples [13,31,32]. It is also observed that the CuO peak
intensities increased for higher calcination temperatures [12] for all
the samples especially for those prepared by the wet impregnation
technique. However, the peak of crystalline CuO was less noticeable
for samples prepared using complex precipitation (not shown here)
compared to the wet impregnated ones. This could be due to the

Table 3
Band gap energies derived from UV-Vis data for the prepared samples compared to
Degussa P25.
Photocatalyst

Band gap energy, eV

TiO2 (Degussa P25)


10CuGT3_30
10CuGT4_30
10CuGT5_30
10CuT3_30
10CuT4_30
10CuT5_30

3.20
2.91
2.9
2.75
2.40
2.85
2.58

Fig. 7. XRD diffractograms for Cu/TiO2 and TiO2 photocatalysts.

high degree of dispersion of the CuO species present on the surface


of the support [30]. As for the peaks of the crystalline form of TiO2,
these remain essentially unchanged.
3.5. SEM
Typical scanning electron micrographs of the catalyst samples
(Fig. 8) depict the particle structures like irregular spherical of size
around 20100 nm. Fig. 8a shows the TiO2 particles used in this
work. The morphologies of particles prepared by the complex
precipitation method show agglomerates formed at an elevated
calcination temperature of 500  C because of sintering (Fig. 8b, c
and d). Fig. 8e, f and g show platelet shape morphologies (indicated
by arrows) of the 10CuT calcined at 300, 400 and 500  C, respectively. These are deposited copper clusters as CuO on the surface of
the support indicating weak interaction of the CuO particles with
TiO2. Although the band gaps for the wet impregnated catalysts,
10CuT, displayed a large reduction, their lesser photocatalytic

Author's personal copy

1658

L.S. Yoong et al. / Energy 34 (2009) 16521661

Fig. 8. SEM micrographs of the Cu/TiO2 and TiO2 photocatalysts.

Author's personal copy

L.S. Yoong et al. / Energy 34 (2009) 16521661

activity could be due to the CuO particles not adequately interacting


with TiO2. However, morphology obtained by the complex
precipitation method did not show any localized metal deposition.
It is believed that the copper clusters were evenly dispersed or
incorporated into the support during the preparation as indicated
by the XRD patterns (Fig. 7b).
3.6. Photocatalytic activity hydrogen production
A few previous studies indicated a strong correlation between
the hydrogen yield and the calcination temperature that controls
the crystallinity, surface area and photocatalytic activity [34].
Photocatalytic activity of Cu-doped TiO2 catalysts calcined at
different temperatures was measured using the experimental
setup shown in Fig. 1. Fig. 9 depicts the effects of calcination
temperature and metal loading on the photocatalytic activity in
terms of hydrogen production. It appears that the hydrogen
production increases with metal loading for the complexprecipitated catalysts. However, the performance of the catalyst deteriorates at 15% loading of the dopant. The highest hydrogen
production was achieved with the catalysts having 10 wt.% dopant.
The wet impregnated catalysts (CuT series) did not show any
denite trend in the effect of metal loading and calcination
temperature on hydrogen production. Only little change in
hydrogen production is observed when the copper loading is
increased from 2% to 15%. However, 10CuT catalysts were selected
for comparison with 10CuGT catalysts in the study. The hydrogen
yields for the samples 10CuGT and 10CuT calcined at 300  C were
8.5 mL h1 and 4.0 mL h1 which are higher than other prepared
samples and also the commercially available titania, viz., Degussa
P25, which exhibits photocatalytic activity of 2.5 mL h1 only in
terms of hydrogen generation.
It will be pertinent to compare the activity of the Cu-doped
catalysts prepared and used in this work with those reported in the
scanty literature available on doped TiO2 photocatalysts. Bandara
et al. [15] prepared CuO-doped titania catalysts and observed the
best performance at a copper loading of 7% in contrast to 10% in our
work. They reported maximum hydrogen generation pf 20 mL/h
using UV-Vis radiation from a 125 W medium pressure mercury
lamp and 5% methanol in water as the sacricial electron donor. So
far the calcination temperature is concerned, they got maximum
hydrogen production with wet impregnated catalyst calcined at
300  C which is the same as in our work. Yasomanee and Bandara
[17] prepared Cu(1)-doped TiO2 catalysts in the form a thin lm on
a glass plate by electrodeposition of copper. They reported
hydrogen generation of 0.150.27 mL only using 0.5 g copperdoped titania over a reaction time of 2 h. Choi and Kang [13] also

300C

Hydrogen evolved, mLh-1

400C

1659

Table 4
Calculated band gap for the nickel-doped catalysts.
Photocatalyst

Band gap energy, eV

TiO2 (Degussa P25)


10NiGT3_30
10NiGT4_30
10NiGT5_30
10NiT3_30
10NiT4_30
10NiT5_30

3.2
2.9
2.9
2.9
2.9
2.9
2.8

used a CuO-impregnated TiO2 (anatase form) catalyst, calcined at


500  C, and reported a hydrogen yield of 290 mL over a 10 h reaction time. They used a much higher concentration of methanol
(50%) than in our work, a larger quantity of catalyst (2 g) and also
used a UV radiation source for the photocatalytic reduction of
water. However, they reported the best catalytic activity at 10%
copper loading as in our work. Gurunathan [34] prepared copperand cadmium-doped TiO2 catalysts and immobilized Rhodopseudomonas capsulata on it to enhance the catalytic activity of the
material. A hydrogen production rate of 1.34 mL/h was achieved at
a catalyst loading of about 0.1 g in 80 mL water. A few sacricial
agents were used and oxalic acid was found to offer the best results
among them. Considering the above results together with the fact
that we used simulated solar radiation only and a lesser concentration of the sacricial agent in our work, the activity of our
catalyst seems to work better, at least in the visible range of the
spectrum.
A set of nickel-doped photocatalysts was prepared by following
the same procedures of precipitation and wet impregnation. The
same dopant loading and calcination temperatures were used.
Characterization of the nickel-doped catalysts was done also
following the same techniques and procedures. The results of
characterization, however, are not presented here. The calculated
band gap on nickel-doping is shown in Table 4 and the hydrogen
generation data are presented in Fig. 10. The reduction in band gap
is less for nickel-doping and the hydrogen generation is about 60%
of that for copper-doped catalyst. It may be noted, however, that
the highest activity of both types of catalysts occurs at 10% dopant
concentration and 300  C calcination temperature. It may be noted
that Sreethawong et al. [14] who worked with NiO-doped titania
catalyst exposed to UV radiation and used methanol as the hole
scavenger also reported a low hydrogen evolution.
The results of our study further indicate that the photocatalytic
activity of the prepared samples decreases with the increased
calcination temperature. This may be attributed to the growth and
agglomeration of the particles and reduced contact area of particles

500C

8
7
6
5
4
3
2
1
0
2CuGT

2CuT

5CuGT

5CuT

10CuGT

10CuT

15CuGT

Metal loading, wt%


Fig. 9. The effect of calcination temperature and metal loading on H2 production.

15CuT

Author's personal copy

1660

L.S. Yoong et al. / Energy 34 (2009) 16521661

H2 evolved, mL/h

5
4
3
2
1

_3
0

_3
0

10
N
iG
T3

10
N
iG
T4

_3
0

3_
30

10
N
iG
T5

iT
10
N

10
N

10
N

iT

4_
/3
0

5_
30
iT

Ti

O
2

Fig. 10. Hydrogen generation with Ni-doped photocatalyst.

[33]. An improvement of the photocatalytic activity depends upon


the efciency of charge capture by the doping centers. A uniform
dispersion of the dopant on TiO2 facilitates migration of electrons
from its conduction band to that of CuO and an agglomeration of
the dopant weakens the interaction of CuOTiO2 and adversely
affect the photocatalytic activity [17]. Such morphological changes
occurring at a higher calcination temperature is also noticeable in
the SEM pictures (Fig. 8). According to Bandara et al. [15], the
crystallinity of CuO is fully formed at 300  C and the crystalline
nature of CuO formation is vital for the photocatalytic hydrogen
production. On the other hand, a higher calcination temperature
not only leads to the formation of larger particles and lesser contact
area of CuO particles and TiO2, but results in a longer distance of
electron transfer and a higher rate of recombination of photogenerated electronhole pair and a reduced rate of hydrogen
evolution as well for both kinds of samples calcined at 500  C [35].
The improvement of the photocatalyst activity with the Cu-doping
may be partly due to the color of the ions that acts as a chromophore, which absorbs light in visible range [26]. This can also be
observed from the DR-UV-Vis spectra of 10CuGT and 10CuT where
the absorption edges shifted to longer wavelength (Fig. 5a and 5b)
side of the spectrum. It is also to be noted that the amount of
methanol (sacricial electron donor) in each of the experiments
was the same (0.5 mL) and the effect of methanol concentration has
not been investigated in this study. Use of a sacricial agent
suppresses recombination of photogenerated electrons and holes
leading to enhanced photocatalytic water reduction. If some
industrial waste material can be identied as the sacricial agent,
a synergistic benet may be expected.

Fig. 11. A schematic mechanism of the photocatalytic reduction of water to hydrogen.

Here e is the value of the concerned measured variable, De is the


error in measurement and xi (i 1, 2.n) are the independent
variables. The above three major independent variables have the
following maximum measurement errors 1%, 5% and 2%,
respectively. Using these values and the observed dependence of
the error on the selected variables, the maximum and probable
errors are calculated as 6.7% and 4.5%, respectively.
3.8. Mechanism of hydrogen production
A simplied mechanism for the photohydrogen production from
water in the presence of methanol is shown in Fig. 11. The incident
radiation ejects electrons from the valence band to the conduction
band, the band gap energy being reduced by copper doping. A
photo-ejected electron reduces water to hydrogen while methanol,
the hole scavenger, gets oxidized [15]. Since the conduction band of
CuO is located below that of TiO2, electron transfer to the
conduction band of CuO is thermodynamically permissible. This
phenomenon becomes important when a part of the incident
radiation has wavelength in the UV range. Hole transfer from the
valence band of TiO2 to that of CuO is also allowable for the same
reason. The combined effects are responsible for the enhanced
hydrogen production.

3.7. Error analysis

4. Conclusion

The performance of a catalyst sample is ultimately measured in


terms of hydrogen production. For a given catalyst composition, the
accuracy of measurement of hydrogen production depends upon
the calcination temperature, the radiant energy ux from the
simulated source of solar radiation, and the accuracy of measurement of gas volume. The maximum and probable coefcients of the
certainty can be expressed by the following equations [36].

The experimental results on hydrogen production for the two


sets of catalysts prepared with different dopant loadings and
under different conditions of calcination show some denite
trends. The 10CuGT photocatalyst prepared by the complex
precipitation method with 10% copper loading and calcined at
300  C showed better activity than the wet impregnation catalyst.
The hydrogen yield is much higher than that achieved with TiO2
alone. The size range of the dopant particles for the 10CuGT
catalyst varied between 20 and 110 nm. The relatively uniform
dispersion of CuO on TiO2 indicated by the SEM pictures and the
full crystallinity conrmed by the XRD analysis are the factors
behind its highest catalytic efciency. The rate of hydrogen
generation under simulated solar radiation compares well with
that reported in the literature using UV radiation. This particular

Demax 


n 
X

 ve ,Dxi
vx 

i1

Dep 

n 
X
ve
i1

vxi

(4)

,Dxi

2

(5)

Author's personal copy

L.S. Yoong et al. / Energy 34 (2009) 16521661

attribute of the catalyst establish its potential for further development with a view to practical application in harnessing solar
energy. Since uniform dispersion of the dopant ensures fast
charge transfer from TiO2 to CuO, it will be worthwhile to modify
the preparation method to this effect. Synthesis of titania by the
solgel method rather than using the Degussa P25 variety may be
a better option since this material can be obtained in essentially
anatase form by suitably selecting the sintering temperature. Also,
the copper salt may be incorporated during the solgel preparation step for uniform dispersion of the dopant in the nal
product. Other potential dopant materials such as Cu2O and CdS
may be tested for their efciency. The catalysts in the form of
powder have been used in this work. The efciency of utilization
of the catalyst may be improved if it is prepared as a thin lm on
a suitable surface. Studies on the effect of pH of the liquid and
alternative sacricial agents such as glycerol, citric acid and oxalic
acid on hydrogen production may be other aspects of future work.
References
[1] Jae S. Photo-catalytic water splitting under visible light with particulate
semiconductors catalyst. Catal Surv Asia 2005;9:21728.
[2] John NA. The multiple role for catalysis in the production of H2. Appl Catal A
Gen 1996;176:15976.
[3] Momirlan M, Veziroglu T. Recent direction of world hydrogen production.
Renew Sust Energ Rev 1999;3:21931.
[4] Tseng P, Lee J, Eriley P. A hydrogen economy: opportunities and challenges.
Energy 2005;30:270320.
[5] Penner SS. Steps towards hydrogen economy. Energy 2006;31:3343.
[6] Guo LJ, Zhao L, Jing DW, Lu YJ, Yang HH, Bai EF. Solar hydrogen production and
its development in China. Energy 2009;. doi:10.1016/j.energy.2009.03.012.
[7] Carp O, Huisman CL, Reller A. Photoinduced reactivity of TiO2. Prog Solid State
Chem 2004;32:33177.
[8] Wad A. Photocatalytic properties of TiO2. Chem Mater 1993;5:2803.
[9] Ha SP, Dong HK, Sun JK, Kyung SL. The photocatalytic activity of 2.5 wt.% Cudoped TiO2 nano powder synthesized by mechanical alloying. J Alloy
Compounds 2005;415:515.
[10] Ni M, Micheal KH, Dennis YCL, Leung KS. A review and recent development in
photocatalytic water-splitting using TiO2 for hydrogen production. Renew Sust
Energ Rev 2007;11:40125.
[11] Tseng IH, Jeffrey CSW. Chemical states of metal-loaded titania in the photoreduction of CO2. Catal Today 2004;97:1139.
[12] Wu NL, Lee MS. Enhanced TiO2 photocatalysis by Cu in hydrogen production
from aqueous methanol solution. Int J Hydrogen Energy 2004;29:16015.
[13] Choi H-J, Kang M. Hydrogen production from methanol/water decomposition
in a liquid photosystem using the anatase structure of Cu loaded TiO2. Int
J Hydrogen Energy 2007;32:38418.
[14] Sreethawong S, Suzuki Y, Yoshikawa S. Photocatalytic evolution of hydrogen
over mesoporous TiO2 supported NiO photocatalyst prepared by single step
solgel process with surfactant template. Int J Hydrogen Energy 2005;
30:105362.
[15] Bandara J, Udawatta CPK, Rajapakse CSK. Highly stable CuO incorporated TiO2
catalyst for photocatalytic hydrogen production from H2O. Photochem
Photobiol Sci 2005;4:85761.

1661

[16] Bamwenda GR, Tsubota S, Nakamura T, Harutta M. Photoassisted hydrogen


production from water ethanol solution: a comparison of activities of AuTiO2
and PtTiO2. J Photochem Photobiol A Chem 1995;89:17789.
[17] Yasomanee JP, Bandara J. Multi-electron storage of photoenergy using Cu2O
TiO2 thin lm photocatalyst. Sol Energ Mat Sol C 2008;92:34852.
[18] Shi D, Yaqing F, Shunhe Z. Photocatalytic conversion of CH4 and CO2 to
oxygenated compound over Cu/CdSTiO2/SiO2 catalyst. Catal Today 2004;
98:5059.
[19] Liu AR, Wang SM, Zhao YR, Zhang Z. Low temperature preparation of nanocrystalline TiO2 photocatalyst with a very large specic surface area. Mater
Chem Phys 2006;99:1314.
[20] Boccuzzi F, Chiorino A, Martra G, Gargano M, Ravasio N, Carrozzini B. Preparation, characterization and activity of Cu/TiO2 catalysts. I. Inuence of the
preparation method on the dispersion of copper in Cu/TiO2. J Catal
1997;165:12939.
[21] Li Z, Bo H, Yao X, Dong W, Yuhan S, Wei H, et al. Comparative study of solgel
hydrothermal and solgel synthesis of titaniasilica composite nanoparticles.
J Solid State Chem 2005;178:1395405.
[22] Yan X, He J, Evans DG, Zhu Y, Duan X. Preparation, characterization and
photocatalytic activity of TiO2 formed from mesoporous precursor. J Porous
Mat 2004;11:1319.
[23] Anpo M, Takeuchi M. The design and development of highly reactive titanium
oxide photocatalysts operating under visible light irradiation. J Catal
2003;216:50516.
[24] Colon G, Maicu M, Hidalgo MC, Navio JA. Cu-doped TiO2 system with
improved photocatalytic activity. Appl Catal B Environ 2006;67:4151.
[25] Akihiko K, Ryo N, Akihide I, Hideki K. Effects of doping of metal cations on
morphology, activity and visible light response of photocatalyst. Chem Phys
2007;339:10410.
[26] Sadhana R, Nidhi D, Mitin KL, Kagne S, Sukumar D. UV and visibly active
photocatalyst for water splitting reaction. Int J Hydrogen Energy 2007;
32:277683.
[27] Diamandescu L, Vasiliu F, Tarabasanu-Mihaila D, Feder M, Vlaicu AM,
Teodorescu CM, et al. Structural and photocatalytic properties of iron and
europium-doped TiO2 nanoparticles obtained under hydrothermal conditions.
Mater Chem Phys 2008;112:14653.
[28] Murphy AB. Band-gap determination from diffuse reectance measurement of
semiconductor lms, and application to photoelectrochemical water-splitting.
Sol Energ Mat Sol C 2007;91:132637.
[29] Slamet HWN, Ezza P, Soleha K, Jarnuzi G. Photocatalytic reduction of CO2 on
copper-doped titania catalysts prepared by improved-impregnation method.
Catal Commun 2005;6:3139.
[30] Yuan Z, Wang Y, Sun Y, Wang J, Bie L. Sunlight-activated AlFeO3/TiO2 photocatalyst. Sci China Ser B Chem 2006;49:6774.
[31] Marc H, Christian B, Jacques L. Synthesis and characterization of copper (II)
hydroxide gels. J SolGel Sci Technol 1996;6:15562.
[32] Albin P, Jurka B, Stanko HTPR. TPO and TPD examinations of Cu0.15Ce0.85O2-y
mixed oxides prepared by co-precipitation by the solgel peroxide route and
by citric acid-assisted synthesis. J Colloid Interface Sci 2005;285:21831.
[33] Singto S, Sorapong P, Yoshikazu S, Susumu Y. Photocatalytic activity of titania
nanocrystals prepared by surfactant-assisted templating methoddeffect of
calcination conditions. Mater Lett 2005;59:295668.
[34] Gurunathan K. Photobiocatalytic production of hydrogen using sensitized
TiO2-MV2 system coupled Rhodopseudomonas capsulata. J Mol Catal A Chem
2000;156:5967.
[35] Yi H, Peng T, Ke D, Dai K, Ling Z, Yan C. Photocatalytic H2 production from
methanol aqueous solution over titania nanoparticles with mesostructures.
Int J Hydrogen Energy 2008;33:6728.
[36] Bujak J. Mathematical modelling of a steam boiler room to research thermal
efciency. Energy 2008;33:177987.

You might also like