2h = 25.34 and 2h = 27.42 correspond to the main
peak of anatase and rutile, respectively. It can be seen
that for all copper modified TiO2 samples, only the characteristic
peaks corresponding to P25 were found (which
consisted of 79% anatase and 21% rutile). CuO diffraction
peak appeared near 2h = 35.6 only on the high
CuO loading samples (P3%), and the peak intensities
increase for higher CuO loading. The disappearance of
the CuO peaks in the low CuO loading samples is due
to the high dispersion of the dopant species. In addition,
the disappearance of the CuO peaks after reduction of
the samples (Fig. 2-d0) indicated that CuO has been successfully
reduced to other species. It can also be inferred
from Fig. 2 that the crystalline form of the TiO2 has not
been changed by loading CuO as the shapes and relative
intensity of the characteristic TiO2 peaks remained
unchanged.
To discuss the interactions between the dispersed
ionic species and the support, the surface structure of
the support should be taken into consideration. According
to Xu et al. [7] only the octahedral vacant sites are
available in the preferentially exposed (0 0 1) plane of
anatase, and the dispersion capacity of CuO on TiO2
(anatase) should be equal to the number of the surface
vacant sites available, i.e., 6.98 Cu2+ nm2. When the
CuO loading exceeds its dispersion capacity, all the vacant
sites have been occupied, and then the existence
of crystalline CuO can be detected by XRD.
Plotted in inset of Fig. 2 is the ratio of the XRD peak
intensities of CuO and TiO2 vs. the loading amounts of
CuO in the samples, from which the dispersion capacity
of CuO on TiO2 P25 is determined to be 2.2 wt% (4.16
Cu2+ nm2). Whereas Xu et al. [7] have investigated that
the dispersion capacity of CuO on TiO2 (anatase) is
about 6.74 Cu2+ nm2. The discrepancy of this dispersion
capacity may be caused by the different TiO2 used
as precursor. Xu et al. [7] used a pure TiO2 anatase,
and in this work, we used TiO2 Degussa P25 which consisted
of 79% anatase and 21% rutile; this may also be
due to the higher of thermal treatment in this work
(500 C) than their work (450 C).
Table 1 also presents the characteristic of CuO/TiO2
catalysts from XRD analysis. The crystalline size of anatase
and rutile were relatively uniform, ranging from 19
to 22 nm and 23 to 31 nm, respectively, whereas the Cu
clusters grew up with the increase of its loading. Addition
of copper in the catalysts leads to the decrease of
anatase content, due to the presence of thermal treatment
step in catalyst preparation.
SEM analysis of the catalysts indicated that addition
of Cu on TiO2 can affect the surface morphology of the
catalyst, however there is no significant effect on the
aggregate sizes. The aggregate size of the catalysts were
relatively uniform, ranging from 0.4 to 1.0 lm. The
EDX/Mapping analysis of Cu on the catalysts demonstrated
that Cu uniformly dispersed on TiO2 surface,
and further increase of copper leads to the increase of
the dot intensities of Cu. In addition, Table 1 lists the
elemental wt% of Cu estimated from EDX and AAS
for comparison. The difference between wt% of Cu obtained
by EDX on different voltage (15 and 22 kV)
and AAS indicated that most Cu was on the surface
of the TiO2.
The DRS patterns of TiO2 P25 and all CuO-loaded
samples are shown in Fig. 3. The figure shows that the
spectra of fresh and reduced copper–titania catalysts
are different from unmodified TiO2 P25. Surface modification
of TiO2 with copper significantly affects the
absorption properties of the catalysts. The absorption
spectra were obtained by analyzing the reflectance measurement
with Kubelka–Munk emission function [5]. It
is noticeable that the absorbance of the copper loaded
samples increase with increasing Cu content. Copper
modified catalyst caused absorption spectra to shift to
the visible region, in the range of 415–700 nm.