270
L.H. Oliveira et al. / Journal of Alloys and Compounds 647 (2015) 265e275
contributing to the conduction mechanism in the matrix. The
captured electrons in conduction band contributes to non-radiative
processes, competing with the energy transfer from matrix to the
excited levels of Pr3þ ions, corroborating to a decrease in the in-
tensity of the 4f/4f emission. The excess of positive charge can be
the degree of structural orderedisorder of the material and
consequently, the number of intermediary energy levels within
band gap [44]. Thus, decrease in the bandgap can be attributed to
defects, such as distortions along CaeO linkages which yield
localized electronic levels in the band gap. Moreover, the presence
4
þ
compensated by intertitals defects, when Ti sites are substituted
of TiO
2
secondary phase can be responsible for the conductive
TiO
3
þ
by Al ions, improving its optical properties. Mazzo et al. [39] also
noticed an improvement in local order-disorder degree by creation
character observed for Ca1-xMg
x
3
powders with x ¼ 0.005 and
2þ
x ¼ 0.01 of Mg ions.
3
þ
of cations or oxygen vacancies with substitution of Eu ions in
both A and B sites of CaTiO matrix. This factor also provided sig-
Ca1-xMgx/2Eu2y/3TiO
3
powders with x ¼ 0.001, 0.002, 0.003 and
3
0.004 presented Egap values of 3.51 eV, 3.55 eV, 3.52 eV and 3.66 eV,
respectively. In the Ca1-xMg Eu TiO powders, when the concen-
tration of Mg and Eu increases from 0.000 to 0.003, Egap values
are practically constant. Absorbance measurements for these
samples suggests that the density of localized states in the band gap
nificant changes in local order-disorder degree around Ca and Ti
clusters.
y
z
3
2þ
3þ
Based on these papers, we believe that symmetry break along
6
[TiO ] and [CaO12] clusters is mainly caused by disorder into the
2þ
3þ
lattice modifiers (Ca atoms) due to distorted [CaO12] clusters and/or
did not change as Mg and Eu ions were added to the CaTiO
3
2
þ
Ca vacancies in Ca1-xMg
x
TiO
3
and Ca1-xMgx/2Eu2y/3TiO
3
powders.
matrix. The x ¼ 0.004 powder presented an increase in the Egap
when compared to the other samples. This behavior can be asso-
ciated to the formation of a solid solution, being in accordance to
the Rietveld refinement analysis.
These species are strongly related to the formation of
ꢃ
0
z
½
CaO ꢄ 4½CaO ꢄ clusters (o ¼ ordered, d ¼ disordered) and V
¼
1
2 d
12 o
Ca
ꢃ
ꢃꢃ
x
Ca
V
þ V þ V species in CaTiO
3
matrix, respectively. These spe-
Ca
Ca
cies are neutral and ionized centers, creating electronehole pairs
ꢃ
ꢂ
(
hole (h )ꢂelectron (e )-polarons) in the Ca1-xMg
x
TiO
3
and Ca1-
3.4. Photoluminescence measurements
x
Mgx/2Eu2y/3TiO powders, contributing to the conduction mecha-
3
nism in the matrix [35].
Since Canham et al. [45] studies, the PL spectroscopy has been
the objective of several studies and it is considered a powerful
technique to understand the carrier recombination and quantita-
tive characterization of crystals and heterointerfaces [36]. The
proposed model by Anicete-Santos et al. [46] considered photo-
luminescent emission recurrent from radiative recombination of
3.3. UVevis absorbance spectroscopy
In this work, the band gap energy (Egap) of the Ca1-xMg
with x ¼ 0.00, 0.005, 0.01 and 0.02 of Mg and Ca1-xMg
x
TiO
TiO
3
2
þ
ꢂ
y
Eu
z
3
ꢃ
(
h )ꢂelectron (e )-polarons in the existing energy levels between
powders with x ¼ 0.000, 0.001, 0.002, 0.003 and 0.004 were
measured by Kubelka-Munk method [40]. The band gap energy is
associated to the electronic transitions, which occurs from the
located states in the conduction band (CB) to the located states in
the valence band (VB). A direct band gap is related to the electronic
transitions that occur from localized states in the same Brillouin
region [35], while the indirect band gap is associated to localized
states in a different Brillouin region. The band gap (Egap) and the
BV and BC, which is associated with the presence of imperfections
or defects in the crystal lattice. To evaluate the effect of Mg doping
in Ca1-xMg
response for all powders was studied under excitation centered in
the host matrix (
¼ 350 nm, room temperature).
PL emission of Ca1-xMg TiO powders (Fig.4(a)) covers the
x 3
TiO prepared by the PPM, the photoluminescence
l
x
3
visible and infrared region of the electromagnetic spectrum. Three
broad bands emissions are observed: one situate at 445 nm (blue
absorption coefficient (a) of a semiconductor oxide are associated
region) associated to CaTiO
3
matrix, a broad band at 625 nm (red
according to the equation (3):
region) ascribed to MgTiO [47] phase and a broad band at 800 nm
3
ꢀ
(infrared region) associated to the presence of rutile TiO [48].
ꢁ
ꢂ
2
n
ahn ¼ C1
R
∞
hn ꢂ Egap
(3)
The obtained results showed that all these systems presented a
PL emission band characteristic of a multiphotonic process, i.e., a
system in which relaxation occurs by several paths involving the
participation of numerous states within the bandgap of the mate-
rial. Thus, the decomposition of these broad bands can be used to
get the information of which electronic transitions group is influ-
where
a
is the linear absorption coefficient of the material, h
photon energy, C
n
is the
1
is a proportional constant, Egap is the optical band
gap and n is a constant associated to the different types of elec-
tronic transitions (n ¼ ½ for direct allowed, n ¼ 2 for indirect
allowed, n ¼ 3/2 for direct forbidden and n ¼ 3 for indirect
forbidden). Egap values were measured considering a direct allowed
electronic transition in equation (3) and the values were evaluated
by the extrapolation of the linear portion of the curve considering
encing the PL response (Fig. 5). The decomposition of the CaTiO
3
emission band was performed using the deconvolution method by
PeakFit [49] Program (version 4.05). The Gaussian function was
successfully used to fit the PL peaks and tuning parameters,
including the peak positions and its corresponding areas.
From deconvolution method, it was verified that the PL emission
x 3
of Ca1-xMg TiO powders are composed by three components; a
violet component with a maximum situated at around 420 nm, a
blue component with a maximum at around 450 nm and the green
component with a maximum at 505 nm. Besides that, PL curves is
centered at the blue region of the electromagnetic spectrum, and
linear regression of r ¼ 0,99. For Ca1-xMg
x 3
TiO powders (Fig. 3(a))
were founded a Egap values of 3.51 (x ¼ 0.000), 2.96 (x ¼ 0.005), 2.89
(
x ¼ 0.01) and 3.44 (x ¼ 0.02), respectively.
The calculated band structures [35,34,33,41] show that the
electronic transition for CaTiO
3 6
powders occurs inside [TiO ]
octahedral clusters since 2p orbitals of oxygen atoms in valence
band and 3d orbitals of the titanium atoms can be associated with
the conduction band. These electronic states are located in the
Ca1-xMg
x
TiO
3
(x ¼ 0.01) powder presented the highest percentage
same Brillouin region. However, for Ca1-xMg
x
TiO
3
powders 3s or-
(74.5%). This observation confirmed that the PL response is directly
bitals states of Mg atoms will be associated to the conduction band
ascribed to localized states in the band gap due to structural defects
[42,43], so a decrease of the optical gap is observed.
x 3
in the Ca1-xMg TiO crystalline structure. The blue emission (more
The different band gap values indicate the existence of various
energetic) indicates the redistribution of energetic states, which is
related to the insert of deep defects between conduction and
valence bands.
intermediary levels between the CB and VB, which is affected by
2
þ
Mg concentration in the A-site of CaTiO
3
. This factor can favor or
MR and Rietveld analyses showed that the introduction of Mg2
þ
inhibit the formation of structural defects, which are able to control