J Mater Sci (2010) 45:2696–2702

DOI 10.1007/s10853-010-4254-5

Photocatalytic reduction of methylene blue by TiO2 nanotube

arrays: effects of TiO2 crystalline phase
Xiongwu Kang • Shaowei Chen

Received: 13 October 2009 / Accepted: 20 January 2010 / Published online: 4 February 2010
Ó The Author(s) 2010. This article is published with open access at Springerlink.com

Abstract TiO2 nanotube arrays were synthesized by these crystalline grains that impeded the dynamics of
anodization of Ti metal sheets followed by thermal electron–hole recombination. These results demonstrate the
annealing at elevated temperatures from 400 to 600 °C. significance of nanoscale engineering in the manipulation
Scanning electron microscopic measurements showed that of oxide photocatalytic performance.
dense arrays of nanotubes were produced with the inner
diameter about 100 nm, wall thickness 35 nm, and length
about 10 lm. X-ray diffraction measurements showed that Introduction
the as-prepared nanotubes were largely amorphous,
whereas thermal annealing led to the formation of well- Photocatalysis of a chemical reaction at a semiconductor
defined anatase crystalline phase. More interestingly, at surface entails oxidation by the photogenerated holes and
470 °C, the brookite crystalline phase also started to reduction by the photogenerated electrons that occur
emerge, which became better defined at 500 °C and dis- simultaneously [1]. Titania, a kind of semiconductor
appeared eventually at higher temperatures, a phenomenon materials with very low cost, chemical inertness and
that has not been observed previously in TiO2 nanotube photostability [2–4], has attracted extensive attention as
arrays prepared by anodization. The impacts of the TiO2 one of the most promising materials in solar cells and
nanocrystalline structure on the photocatalytic activity photocatalysis [5–7]. The earliest work of photocatalysis of
were then examined by using the reduction of methylene titania was reported in 1921 by Renz [8], where it was
blue in water as an illustrating example. Upon exposure to found that titania turned from white to a dark color under
UV lights, the visible absorption profiles of methylene blue sunlight illumination in the presence of an organic com-
exhibited apparent diminishment. Based on these spectro- pound such as glycerol. Photocatalytic activity of TiO2 was
photometric measurements, the corresponding pseudo-first- also demonstrated by Fujishima and Honda [9] where they
order rate constant was estimated, and the sample ther- used TiO2 for water splitting, an important process for solar
mally annealed at 500 °C was found to exhibit the highest energy conversion. In addition, Frank and Bard [10, 11]
activity. The strong correlation between the TiO2 crystal- showed that the photocatalytic activity of TiO2 might be
line characteristics and photocatalytic performance sug- exploited for the decomposition of pollutants under photo
gests that the synergistic coupling of the anatase and irradiation. In these reactions, the fundamental mechanisms
brookite crystalline domains led to effective charge sepa- [12–14] entail the rapid excitation of TiO2 valence elec-
ration upon photoirradiation and hence improved photo- trons to the conduction band by absorbing photons with
catalytic activity, most probably as a consequence of the energy larger than the TiO2 bandgap (*3.0 eV), where the
vectorial displacement at the nanoscale junctions between photogenerated electrons (conduction band) and holes
(valence band) are involved in varied redox reactions.
Nevertheless, the efficiency of TiO2 photocatalysis is
X. Kang
S. Chen (&)
generally rather low. Therefore, substantial efforts have
Department of Chemistry and Biochemistry, University
of California, 1156 High Street, Santa Cruz, CA 95064, USA been devoted to engineering the TiO2 crystalline and
e-mail: schen@chemistry.ucsc.edu morphological structures for the improvement of the

123
J Mater Sci (2010) 45:2696–2702 2697

photocatalytic activity. One effective approach is to fabri- temperatures (550–600 °C), the brookite phase disappeared
cate nanotubule arrays of TiO2 by taking advantage of the and only the anatase phase was found within the TiO2
high surface-to-volume ratio, the ready accessibility of the nanotube arrays. Overall, the photocatalytic activity of
TiO2 surface to reactants and easy release of photoreaction these thermally annealed TiO2 nanotube arrays, as mani-
products. A number of synthetic protocols have been fested in the photoreduction of methylene blue, displayed a
reported in the literature for the fabrication of TiO2 nano- marked improvement as compared to that of the amorphous
tubes, including electrochemical anodic oxidation of tita- as-prepared sample, and reached the maximum with sam-
nium metal sheets [15], as well as template-based atomic ples that were thermally annealed at 500 °C (with a binary
layer deposition [16], sol–gel [17], or liquid phase depo- mixture of anatase and brookite phases), suggesting the
sition [18–21] method. Among these, Grimes and impacts of the synergistic interactions between the ana-
coworkers [15] reported an effective route in 2001 toward tase and brookite crystalline domains on the photocata-
the fabrication of titania nanotube arrays by anodization of lytic activity. Such a strong correlation demonstrates the
a titanium foil in fluoride-based electrolyte solutions by significance of nanoscale engineering in the manipulation
using fluoride ions as effective etchants. It has been found and optimization of the TiO2 structures, properties and
that the nanotube morphology [22], length and pore size functions.
[23], as well as wall thickness [24] might be readily con-
trolled by varied experimental parameters such as elec-
trolyte composition, solution pH, as well as electrode Experimental section
voltage [25].
Furthermore, postsynthesis thermal annealing offers a Methylene blue (ACROS), ethylene glycol (99?%,
powerful route toward the manipulation of the TiO2 crys- Aldrich) and ammonium fluoride (NH4F, 98?%, extra
talline characteristics, in particular, transformation from pure, ACROS) were used without further treatment. Tita-
one crystalline phase to another as well as control of the nium foils (0.25 mm thick, 99.8% purity (metal basis))
size of the nanocrystalline grains [26]. Note that titania of were purchased from Aldrich. Before the experiments, the
both anatase and rutile crystalline phases has been found to titanium sheets were degreased by sonication in acetone,
be photocatalytically active, although it is generally ethanol, and Nanopure water successively and then dried in
accepted that anatase TiO2 exhibits a better performance, nitrogen. TiO2 nanotube arrays were prepared by using a
largely because of the somewhat higher Fermi level and the two-electrode setup in which a titanium foil was used as the
lower recombination rate of its photogenerated electrons sacrificial anode and a Pt wire as the cathode, along with an
and holes than those of the rutile phase [27–30]. In addi- electrolyte consisting of 0.25% NH4F in ethylene glycol.
tion, the coexistence of both anatase and brookite or rutile All the electrochemical reactions were carried out at room
crystalline phases has also been found to lead to further temperature by using a DC regulated power supply (B&K
improvement of the photocatalytic performance, as a con- Precision Corp., Model 1623A) at the constant voltage of
sequence of the vectorial displacements at the nanoscale 60 V. The nanotube arrays were then subject to thermal
junctions between two semiconducting domains that fur- annealing in air at varied temperatures in a muffle furnace
ther impede the dynamics of electron–hole recombination for 6 h.
[31–37]. However, in these earlier studies, the TiO2 sam- The structures of the TiO2 nanotubes were then exam-
ples with a mixture of the crystalline phases were mostly ined by SEM measurements with an ISI WB-6 scanning
prepared by the sol–gel methods in the form of nanopar- electron microscope operated at 5 kV. The crystalline
ticles, powders, or thin films, and it has remained a chal- characteristics were evaluated by powder X-ray diffraction
lenge to identify optimal experimental conditions for the (XRD) measurements, where the nanotubes were scratched
production of TiO2 nanotubes with significant brookite off from the titanium metal sheets and deposited onto an Al
contents. plate, with a Rigaku Mini-flex Powder Diffractometer
In this study, we demonstrate for the first time that using Cu-Ka radiation with a Ni filter (k = 0.154059 nm at
brookite crystalline characteristics may also be produced in 30 kV and 15 mA). The average crystalline grain size was
anodized TiO2 nanotube arrays by a deliberate manipula- calculated from the XRD line broadening of the TiO2
tion of the thermal annealing conditions. The results show anatase (101) or brookite (121) diffraction peaks by using
that upon thermal annealing at 400–600 °C, the nanotube the Debye–Scherrer equation (in the calculations, the
arrays exhibited clear anatase crystalline characteristics, in instrumental broadening was taken into account, which was
comparison with the largely amorphous structure of the calibrated with a standard Si sample [26, 38]).
as-produced samples. More interestingly, at 470 °C a second The photocatalytic reduction of methylene blue by TiO2
phase of brookite crystalline features also started to appear, nanotubes was monitored spectrophotometrically with an
which became much better defined at 500 °C. At higher ATI Unicam UV4 spectrometer by using a 1-cm quartz

123
2698 J Mater Sci (2010) 45:2696–2702

cuvette with a resolution of 2 nm. Prior to exposure to UV

lights, the TiO2 nanotubes (1 cm 9 1 cm) were immersed
into the methylene blue solution (concentration 26 lM in
water) for at least 40 min in the dark in a nitrogen atmo-
sphere until no further decrease of the absorption intensity
of methylene blue was observed. Optical absorption pro-
files were then acquired at different exposure times to the
UV lights (365 nm, 6 W). Note that the solution was
blanketed with a nitrogen atmosphere during the entire
experimental procedure.

Results and discussion

The surface morphology of the TiO2 nanotube arrays was

first characterized by scanning electron microscopic (SEM)
measurements. Figure 1 depicts two representative SEM
micrographs of a TiO2 nanotube array that was prepared at
an anodic voltage of 60 V and then thermally annealed at
550 °C for 6 h. From the top view (panel a), one can see
that a very dense array of rather uniform tubule structures
was formed, with a tube density of ca. 7.4 9 1011 cm-2,
and the inner diameter of the tube was found to be around
100 nm, and tube thickness ca. 35 nm. From the side view
(panel b), the nanotube exterior surface was rather smooth
and the length was about 10 lm.
Whereas variation of the thermal annealing temperature
did not lead to obvious differences in terms of nanotube
dimensions and surface morphologies, the crystalline
details of the nanotube arrays were found to vary sub-
Fig. 1 Representative SEM micrographs of TiO2 nanotube arrays
stantially, as manifested in powder X-ray diffraction that were prepared by electroanodization at 60 V followed by thermal
(XRD) measurements. Figure 2 shows the XRD patterns of annealing at 550 °C for 6 h. Scale bars are 200 nm in both a top view
the TiO2 nanotube arrays thermally annealed at different and b side view
temperatures from 400 to 600 °C, along with that of the as-
produced sample. In comparison to the standard XRD as manifested by the peaks (marked with ‘‘B’’) at 31.2° and
profiles for anatase, brookite, and rutile TiO2, it can be seen 36.1°, which were ascribed to the brookite (121) and (012)
that the as-prepared sample exhibited only a featureless crystalline lattices, respectively. In other words, within the
profile (curve a), suggesting that the nanotubes were mostly narrow temperature range of 470 to 500 °C, the TiO2
amorphous (the several sharp features marked with ‘‘Al’’ nanotube arrays exhibited a binary mixture of anatase and
arose from the aluminum substrate). In contrast, thermal brookite crystalline phases. To the best of our knowledge,
annealing at elevated temperatures led to effective trans- such an observation has not been reported previously with
formation of the nanotubes into well-defined crystalline TiO2 nanotubes prepared by electroanodization [25]. It
structures. Specifically, for the nanotube arrays that were should be noted that in most early studies where TiO2
thermally annealed at 400–430 °C (curves b and c), a series exhibited a mixture of crystalline phases, the titania sam-
of diffraction peaks appeared (marked with ‘‘@’’), which ples were mostly thin films, powders, or nanoparticles that
are all consistent with the diffraction features of anatase were prepared by sol–gel methods [31–37]. For instance,
TiO2 (JCPDS 75-1537): (101), 25.7°; (200), 47.9°; (105), Junin et al. [33] prepared TiO2 powders by a sol–gel
53.7°; (211), 55.0°; (213), 61.9°; (204), 62.8°; (116), 68.3°; method using titanium tetrachloride and titanium tetraiso-
(220), 69.7°; and (301), 75.2°. When the annealing tem- propoxide as the precursors. A binary mixture of anatase
perature was increased to 470 °C (curve d), in addition to and brookite crystalline phases was found at calcination
the anatase diffraction features, several diffraction peaks temperatures between 400 and 500 °C.
characteristic of the brookite phase also started to appear, Interestingly, with the further increase of the annealing
and they became much better defined at 500 °C (curve e), temperature to 550 and 600 °C, the brookite diffraction

123
J Mater Sci (2010) 45:2696–2702 2699

Fig. 2 XRD patterns of a the

as-prepared TiO2 nanotube Al Al
arrays as well as those that
underwent thermal annealing at
different temperatures:
b 400 °C, c 430 °C, d 470 °C
e 500 °C, f 550 °C, and Al
g 600 °C. Al denotes diffraction Al
features from the aluminum
substrate, ‘‘@’’ refers to @
diffraction peaks of anatase @ @ @ @
@
TiO2, whereas B marks those (g)
of brookite TiO2
(f)
B B
(e)

(d)

(c)

(b)

(a)

20 30 40 50 60 70 80
2θ (degree)

features disappeared and the sample again exhibited a pure crystalline domains that occurred at 470–500 °C (based on
anatase phase, suggesting the back conversion of brookite the brookite (121) diffraction peak, the brookite grains
into anatase rather than further transformation into rutile. were estimated to be about 53.8 nm at 470 °C and 46.2 nm
This indicates that for the TiO2 nanotube arrays prepared at 500 °C). At temperatures higher than 500 °C, the
above, the transformation between anatase and brookite brookite phase disappeared, and the anatase crystalline
was reversible at temperatures up to 600 °C, possibly grains became slightly larger to 29.6 nm at 550 °C and
because of the low activation energy (11.9 kJ/mol [26]) for 25.7 nm at 600 °C [39]. The small anatase crystalline
the transformation between these two phases. grains (Table 1) might also explain why no rutile feature
The grain size of the crystalline domains in the TiO2 was observed even at 600 °C, as the critical size for the
nanotubes was then evaluated quantitatively by the Debye– transformation into rutile has been found to be [35 nm
Scherrer equation, D = Kk/bcosh, where D is the diameter [26].
of the crystalline grain, K = 0.9, k(Cu-Ka) = 1.54059 Å, The effects of the evolution of the anatase nanocrystal-
and b is the full width at half maximum of the selected line grains on the nanotube photocatalytic activity were
diffraction peaks. Based on the anatase (101) diffraction then examined by using the reduction of methylene blue as
peaks, the average size of the anatase crystalline domains the illustrating example. As methylene blue absorbs very
at different annealing temperature was then estimated and strongly in the visible range with two well-defined
listed in Table 1. It can be seen that the variation of the absorption peaks at 612 and 665 nm and the reduced form is
anatase grain size with annealing temperature is relatively colorless, the photocatalytic reduction of methylene blue by
small from 400 to 600 °C. For instance, the anatase grains TiO2 may then be monitored by the spectrophotometric
decreased somewhat from 27.4 nm at 400 °C to 21.3 nm at method. Mechanistically, it is generally proposed that upon
500 °C, possibly because of the formation of the brookite UV illumination, electrons are excited from the TiO2
valence band to the conduction band. The photogenerated
Table 1 Average grain size of the anatase (DA) and brookite (DB) electrons may then be exploited for reduction of methylene
phases in TiO2 nanotube arrays that were thermally annealed at dif- blue. As oxygen is an effective electron scavenger, it is
ferent temperatures (T) important that the reaction media is protected with an inert
T (°C) 400 430 470 500 550 600 atmosphere of nitrogen. Figure 3a depicts the optical
a
absorption profiles of methylene blue after the solution was
DA (nm) 27.4 22.2 25.0 21.3 29.6 25.7
exposed to UV photoirradiation for varied periods of time in
DB (nm)b 53.8 46.2
the presence of a TiO2 nanotube array that was annealed at
a
Based on anatase (101) diffraction peak 500 °C for 6 h. It can be seen that the absorption fea-
b
Based on brookite (121) diffraction peak tures diminish rather quickly upon UV photoirradiation,

123
2700 J Mater Sci (2010) 45:2696–2702

1.8 e1
e1
(a)
1.6 0 min

Peak Absorbance
e0
20 min
40 min e-1 e0
1.4
80 min
e-2
100 min
1.2

Peak Absorb ance

120 min e-3
Absorbance

140 min e-1

e-4
1.0 0 30 60 90 120 150
t (min)
0.8
e-2 as-prepared
0.6 400 oC
430 oC
0.4 470 oC
e-3 500 oC
550 oC
0.2
600 oC
0.0 e-4
500 550 600 650 700 750 800 850 0 50 100 150 200 250 300

λ (nm) t (min)

0.025
Fig. 3 Visible absorption profiles of a methylene blue solution (b)
(26 lM in water) after being exposed to UV photoirradiation for
varied periods of time (shown as figure legends) in the presence of a 0.025
TiO2 nanotube array that was thermally annealed at 500 °C for 6 h.
Inset depicts the semilog plot of the variation of the intensity of the 0.020
absorption peak at 665 nm with UV exposure time. Symbols are

k (min-1)
experimental data and line is linear regression
0.015

0.010
suggesting the effective photocatalytic activity of TiO2 in
the reduction of methylene blue. Panel (b) shows the 0.005
semilog plots of the decay of the absorption intensity at
665 nm with time. The linearity is consistent with the first- 0.000
as-prepared 400 °C 430 °C 470 °C 500 °C 550 °C 600 °C
order reaction of methylene blue reduction [40], and from
linear regression the pseudo-first-order rate constant was Fig. 4 a Semilog plots of the variation of the intensity of the
estimated to be 2.42 9 10-2 min-1. absorption peak at 665 nm with UV exposure time for the photocat-
The photocatalytic activity of the TiO2 nanotubes alytic reduction of methylene blue by TiO2 nanotubes that were
thermally annealed at different temperatures (shown as figure
annealed at other temperatures was evaluated in a similar legends). The experimental conditions were the same as those in
fashion. Figure 4a depicts the decay profiles of methylene Fig. 3. Symbols were experimental data and lines were linear
blue absorption at 665 nm with TiO2 nanotubes that were regressions, from which the pseudo-first-order reaction rate constants
thermally annealed at different temperatures, along with (k) were evaluated, as depicted in panel (b)
that of the as-prepared nanotubes. It can be seen that all
annealed samples exhibited apparent enhancement of the
photocatalytic activity in the reduction of methylene blue as 500 °C, which coincided with the formation of a binary
compared to the as-prepared sample. In addition, among the mixture of anatase and brookite nanocrystalline domains
series, the diminishment of the methylene blue absorption [41, 42], as manifested in Fig. 2. This is most probably
feature was the fastest for the sample annealed at 500 °C. because of the vectorial displacements of the photogener-
Panel (b) shows the pseudo-first-order rate constants (k) that ated electrons at nanoscale junctions between these two
were estimated from linear regressions of the dynamics semiconducting phases that retard the charge recombination
curves in panel (a). It can be seen that for the as-prepared process and hence enhance the charge separation efficiency
sample, k is 2.30 9 10-3 min-1. It increased markedly to [43, 44]. In fact, enhanced photocatalytic performance of
1.00 9 10-2 min-1 and 1.24 9 10-2 min-1 with TiO2 TiO2 with a binary mixture of anatase and brookite crys-
nanotubes thermally annealed at 400 and 430 °C, respec- talline phases, as compared to that of pure anatase, has been
tively. At higher temperatures, the rate constant increased reported in a number of studies [44–47].
further to 1.70 9 10-2 min-1 at 470 °C and 2.42 9 Nevertheless, one might also argue that the variation of
10-2 min-1 at 500 °C. Further increase of the thermal the photocatalytic performance is a consequence of the
annealing temperature actually saw a decrease of the reac- effective oxide surface area that reaches the maximum at
tion rate constant to 9.69 9 10-3 min-1 at 550 °C and 500 °C. However, this is unlikely because an increase of
1.02 9 10-2 at 600 °C. That is, the maximal photocatalytic calcination temperature typically leads to a monotonic
performance was observed with TiO2 nanotubes annealed at decrease of the TiO2 surface area [33]. Additionally, on

123
J Mater Sci (2010) 45:2696–2702 2701

the basis of Raman spectroscopic measurements (not 6. Bozzi A, Yuranova T, Guasaquillo I, Laub D, Kiwi J (2005)
shown), the impact of TiO2 surface functionalities on the J Photochem Photobiol 174:156
7. Ferguson MA, Hoffmann MR, Hering JG (2005) Environ Sci
photocatalytic activity appears to be minimal, as no Technol 39:1880
apparent difference of the Raman spectra was observed 8. Renz C (1921) Helvet Chim Acta 4:961
among the nanotube arrays thermally annealed at different 9. Fujishima A, Honda K (1972) Nature 238:37
temperatures (400–600 °C) [48]. 10. Frank SN, Bard AJ (1977) J Am Chem Soc 99:303
11. Frank SN, Bard AJ (1977) J Phys Chem 81:1484
12. Goodeve CF, Kitchener JA (1938) Trans Faraday Soc 34:570
13. Goodeve CF, Kitchener JA (1938) Trans Faraday Soc 34:902
Conclusion 14. Kennedy DR, Ritchie M, Mackenzie J (1958) Trans Faraday Soc
54:119
15. Gong D, Grimes CA, Varghese OK, Hu WC, Singh RS, Chen Z,
TiO2 nanotube arrays were prepared by anodization of Ti Dickey EC (2001) J Mater Res 16:3331
metal sheets and characterized by SEM and XRD mea- 16. Sander MS, Cote MJ, Gu W, Kile BM, Tripp CP (2004) Adv
surements. SEM study showed that the nanotubes formed a Mater 16:2052
rather dense array, and XRD measurements indicated that 17. Maiyalagan T, Viswanathan B, Varadaraju UV (2006) Bull Mater
Sci 29:705
whereas the as-prepared samples were largely amorphous, 18. Cochran RE, Shyue JJ, Padture NP (2007) Acta Mater 55:3007
thermal annealing at elevated temperatures (400–600 °C) 19. Yamanaka S, Hamaguchi T, Muta H, Kurosaki K, Uno M (2004)
led to the effective formation of crystalline nanotubes with J Alloys Compd 373:312
clear anatase features. Interestingly, at 470 °C, a second 20. Hsu MC, Leu IC, Sun YM, Hon MH (2005) J Cryst Growth
285:642
crystalline phase of brookite features also started to 21. Hamaguchi T, Uno M, Yamanaka S (2005) J Photochem Pho-
emerge, which became better defined at 500 °C. The tobiol 173:99
presence of a binary mixture of anatase and brookite 22. Mor GK, Varghese OK, Paulose M, Mukherjee N, Grimes CA
crystalline phases within this narrow range of annealing (2003) J Mater Res 18:2588
23. Cai QY, Paulose M, Varghese OK, Grimes CA (2005) J Mater
temperature was found to coincide with the maximal Res 20:230
photocatalytic activity of the TiO2 nanotubes in the 24. Mor GK, Shankar K, Paulose M, Varghese OK, Grimes CA
reduction of methylene blue. This is accounted for by the (2005) Nano Lett 5:191
vectorial displacements of the photogenerated electrons at 25. Paulose M, Shankar K, Yoriya S, Prakasam HE, Varghese OK,
Mor GK, Latempa TA, Fitzgerald A, Grimes CA (2006) J Phys
nanoscale junctions between these two semiconducting Chem B 110:16179
phases that retard the charge recombination process and 26. Zhang HZ, Banfield JF (2000) J Phys Chem B 104:3481
hence enhance the charge separation efficiency. This study 27. Mclintock IS, Ritchie M (1965) Trans Faraday Soc 61:1007
demonstrates the fundamental significance of nanoscale 28. Fox MA, Dulay MT (1993) Chem Rev 93:341
29. Brown JD, Williamson DL, Nozik AJ (1985) J Phys Chem
engineering in the manipulation of the properties and 89:3076
functions of nanoscale materials. 30. Maruska HP, Ghosh AK (1978) Sol Energy 20:443
31. Zheng RB, Meng XW, Tang FQ (2009) Appl Surf Sci 255:5989
Acknowledgements This work was supported in part by the 32. Hao HY, Zhang JL (2009) Mater Lett 63:106
National Science Foundation (DMR-0804049) and ACS-PRF (49137- 33. Junin C, Thanachayanont C, Euvananont C, Inpor K, Lim-
ND10). thongkul P (2008) Eur J Inorg Chem 974
34. Tian GH, Fu HG, Jing LQ, Xin BF, Pan K (2008) J Phys Chem C
Open Access This article is distributed under the terms of the 112:3083
Creative Commons Attribution Noncommercial License which per- 35. Tseng YH, Kuo CS, Huang CH, Li YY, Chou PW, Cheng CL,
mits any noncommercial use, distribution, and reproduction in any Wong MS (2006) Nanotechnology 17:2490
medium, provided the original author(s) and source are credited. 36. Hou XH, Choy KL (2005) Mater Sci Eng C 25:669
37. Watson S, Beydoun D, Scott J, Amal R (2004) J Nanopart Res
6:193
38. Chen W, Ghosh D, Chen SW (2008) J Mater Sci 43:5291. doi:
10.1007/s10853-008-2792-x
References 39. Mor GK, Varghese OK, Paulose M, Shankar K, Grimes CA
(2006) Sol Energy Mater Sol Cells 90:2011
1. Fujishima A, Zhang XT, Tryk DA (2008) Surf Sci Rep 63:515 40. Wu XM, Ling YH, Liu L, Huang ZH (2009) J Electrochem Soc
2. Serpone N, Pelizzetti E (1989) Photocatalysis: fundamentals and 156:K65
applications. Wiley, New York 41. Di Paola A, Cufalo G, Addamo M, Ellardita MB, Campostrini R,
3. Schiavello M (1985) Photoelectrochemistry, photocatalysis, and Ischia M, Ceccato R, Palmisano L (2008) Colloids Surf A 317:366
photoreactors: fundamentals and developments. D. Reidel Pub. 42. Zhao B, Chen F, Huang QW, Zhang JL (2009) Chem Commun
Co.; Sold and distributed in the USA and Canada by Kluwer 5115
Academic, Dordrecht 43. Yan MC, Chen F, Zhang JL, Anpo M (2005) J Phys Chem B
4. Linsebigler AL, Lu GQ, Yates JT (1995) Chem Rev 95:735 109:8673
5. Bozzi A, Yuranova T, Kiwi J (2005) J Photochem Photobiol 44. Ozawa T, Iwasaki M, Tada H, Akita T, Tanaka K, Ito S (2005)
172:27 J Colloid Interf Sci 281:510

123
2702 J Mater Sci (2010) 45:2696–2702

45. Ardizzone S, Bianchi CL, Cappelletti G, Gialanella S, Pirola C, 47. Yu JG, Yu JC, Leung MKP, Ho WK, Cheng B, Zhao XJ, Zhao JC
Ragaini V (2007) J Phys Chem C 111:13222 (2003) J Catal 217:69
46. Yu JC, Yu JG, Zhang LZ, Ho WK (2002) J Photochem Photobiol 48. Zhang YH, Chan CK, Porter JF, Guo W (1998) J Mater Res
A 148:263 13:2602

123