The role of oxygen vacancy in the photoluminescence property at room temperature of the CaTiO3 Juliana Milanez, Alberthmeiry T. de Figueiredo, Sergio de Lazaro, Valeria M. Longo, Rafael Erlo et al. Citation: J. Appl. Phys. 106, 043526 (2009); doi: 10.1063/1.3190524 View online: http://dx.doi.org/10.1063/1.3190524 View Table of Contents: http://jap.aip.org/resource/1/JAPIAU/v106/i4 Published by the AIP Publishing LLC. Additional information on J. Appl. Phys. Journal Homepage: http://jap.aip.org/ Journal Information: http://jap.aip.org/about/about_the_journal Top downloads: http://jap.aip.org/features/most_downloaded Information for Authors: http://jap.aip.org/authors Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions http://jap.aip.org/?ver=pdfcov http://oasc12039.247realmedia.com/RealMedia/ads/click_lx.ads/www.aip.org/pt/adcenter/pdfcover_test/L-37/932441298/x01/AIP-PT/JAP_CoverPg_0513/AAIDBI_ad.jpg/6c527a6a7131454a5049734141754f37?x http://jap.aip.org/search?sortby=newestdate&q=&searchzone=2&searchtype=searchin&faceted=faceted&key=AIP_ALL&possible1=Juliana Milanez&possible1zone=author&alias=&displayid=AIP&ver=pdfcov http://jap.aip.org/search?sortby=newestdate&q=&searchzone=2&searchtype=searchin&faceted=faceted&key=AIP_ALL&possible1=Alberthmeiry T. de Figueiredo&possible1zone=author&alias=&displayid=AIP&ver=pdfcov http://jap.aip.org/search?sortby=newestdate&q=&searchzone=2&searchtype=searchin&faceted=faceted&key=AIP_ALL&possible1=Sergio de Lazaro&possible1zone=author&alias=&displayid=AIP&ver=pdfcov http://jap.aip.org/search?sortby=newestdate&q=&searchzone=2&searchtype=searchin&faceted=faceted&key=AIP_ALL&possible1=Valeria M. Longo&possible1zone=author&alias=&displayid=AIP&ver=pdfcov http://jap.aip.org/search?sortby=newestdate&q=&searchzone=2&searchtype=searchin&faceted=faceted&key=AIP_ALL&possible1=Rafael Erlo&possible1zone=author&alias=&displayid=AIP&ver=pdfcov http://jap.aip.org/?ver=pdfcov http://link.aip.org/link/doi/10.1063/1.3190524?ver=pdfcov http://jap.aip.org/resource/1/JAPIAU/v106/i4?ver=pdfcov http://www.aip.org/?ver=pdfcov http://jap.aip.org/?ver=pdfcov http://jap.aip.org/about/about_the_journal?ver=pdfcov http://jap.aip.org/features/most_downloaded?ver=pdfcov http://jap.aip.org/authors?ver=pdfcov The role of oxygen vacancy in the photoluminescence property at room temperature of the CaTiO3 Juliana Milanez,1,a� Alberthmeiry T. de Figueiredo,2 Sergio de Lazaro,3 Valeria M. Longo,2 Rafael Erlo,1 Valmor R. Mastelaro,4 Roberto W. A. Franco,5 Elson Longo,2 and José A. Varela2 1Departamento de Química, Laboratório Interdisciplinar de Eletroquímica e Cerâmica, Universidade Federal de São Carlos, P.O. Box 676, 13565-905 São Carlos, São Paulo, Brazil 2Laboratório Interdisciplinar de Eletroquímica e Cerâmica, Instituto de Química, Universidade Estadual Paulista, P.O. Box 355, 14801-970 Araraquara, São Paulo, Brazil 3Departamento de Química, Universidade Estadual de Ponta Grossa, 84030-900 Ponta Grossa, Paraná, Brazil 4Instituto de Física de São Carlos, Universidade de São Paulo, P.O. Box 369, 13560-970 São Carlos, São Paulo, Brazil 5Centro de Ciência e Tecnologia, Universidade Estadual do Norte Fluminesce, 28013-602 Campos dos Goytacazes, Rio de Janeiro, Brazil �Received 25 February 2009; accepted 1 July 2009; published online 31 August 2009� In this paper, electron paramagnetic resonance, photoluminescence �PL� emission, and quantum mechanical calculations were used to observe and understand the structural order-disorder of CaTiO3, paying special attention to the role of oxygen vacancy. The PL phenomenon at room temperature of CaTiO3 is directly influenced by the presence of oxygen vacancies that yield structural order-disorder. These oxygen vacancies bonded at Ti and/or Ca induce new electronic states inside the band gap. Ordered and disordered CaTiO3 was obtained by the polymeric precursor method. © 2009 American Institute of Physics. �DOI: 10.1063/1.3190524� I. INTRODUCTION Advances in photoluminescence �PL� studies show that titanate compounds with a structural disorder present broad and intense PL at room temperature.1–6 In particular, much work has been concentrated on CaTiO3 �CT� luminescence.7–11 These studies have shown that the broad PL emission band covering a large part of the visible spectra of the structurally disordered compounds is very similar to nanocrystalline PL emission. The PL phenomenon at room temperature occurs due to a structural disorder of the system. However, the system can- not be fully disordered owing to a minimal order in the struc- ture, i.e., there is an order-disorder rate, which favors the PL phenomenon in the system. PL is a powerful probe of certain short-range order aspects in the range of 2–5 Å and in a medium range 5–20 Å, e.g., clusters where the degree of local order is such that structurally inequitable sites can be distinguished because their different types of electronic tran- sitions are linked to a specific structural arrangement. The structural order-disorder of perovskites �especially titanates� has been followed by means of x-ray absorption near-edge spectroscopy �XANES�.12–14 XANES is one of the most suitable techniques to provide information about the atomic environment as it is unique to each element in mate- rials. Ti K-edge XANES results demonstrated that before reaching complete structural order, the cationic environments had fivefold MO5 and sixfold MO6 coordinations �M =Ti�.11,14 In addition, Asokan et al.15 reported changes in the local environment of Ca sites in perovskite structures based on Ca K- and L3,2-XANES spectra. These changes were related to electronic properties derived from those XANES spectra. Similarly, de Lazaro et al.16 reported different coordination modes for calcium by Ca K-edge XANES results. These ex- perimental observations suggest that PL emission is related to �TiO5� and �CaO11� concentrations; moreover, local struc- tural order-disorder depends on both the modifier lattice and the former lattice. Several authors reported structural disorder and quantum mechanics calculation17,18 using periodic calculations DFT/ B3LYP to relate the structural disorder of titanate to room temperature PL properties. The applied methodology is based on atomic displacement to study the charge transfer between �TiO5–TiO6� clusters and electronic levels inside the band gap. In this paper, electron paramagnetic resonance �EPR�, PL emission, and quantum mechanical calculations are used to observe and understand the structural order-disorder, pay- ing special attention to the role of oxygen vacancy. In addi- tion, equations describing the interaction among clusters yielding PL emission in CT are proposed. The use of cluster concepts offers a simple scheme that facilitates an under- standing of the effect of structural deformations on the elec- tronic structure indeed because the most intense PL emission is evidenced in CT samples containing a certain structural order-disorder in the lattice. II. EXPERIMENTAL CT ordered and disordered powders were prepared by the polymeric precursor method and were kept for 20 h at a�Author to whom correspondence should be addressed. FAX: �633518214. Electronic mail: jumilanez@liec.ufscar.br. JOURNAL OF APPLIED PHYSICS 106, 043526 �2009� 0021-8979/2009/106�4�/043526/7/$25.00 © 2009 American Institute of Physics106, 043526-1 Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions http://dx.doi.org/10.1063/1.3190524 http://dx.doi.org/10.1063/1.3190524 http://dx.doi.org/10.1063/1.3190524 300 °C to eliminate any remaining carbon.19 CT was then annealed at 400, 500, or 600 °C for 2 h; CT annealed at 600 °C is fully ordered. The spectral dependence of the op- tical absorbance for ordered �crystalline� and disordered �amorphous� materials was taken at room temperature in the total reflection mode using Cary 5G equipment. The PL mea- surements were collected using 350.7 nm exciting wave- lengths of a krypton ion laser �Coherent Innova� with the nominal output power of the laser kept at 200 mW. The monochromator slit width used was 200 �m. EPR spectra were recorded on a Varian E-9 Spectrometer operating at x-band �9 GHz� and operating at a microwave power of 0.5 mW and modulation frequency of 100 kHz. The g-factor was referenced with respect to MgO:Cr3+ �g=1.9797� as the ex- ternal standard. Spectra were evaluated using the SIMFONIA ® program. Calcium and titanium XANES spectra were col- lected using the D04B-XAS1 beamline. The storage ring was operated at 1.36 GeV and around 160 mA. XANES spectra were collected at the Ca and Ti K-edge �4205 and 4966 eV, respectively� in a transmission mode using a Si�111� channel-cut monochromator. For comparison, all spectra were background removed and normalized using as unity the first extended x-ray absorption fine structure. All experimen- tal measurements were performed at room temperature. III. COMPUTATIONAL DETAILS Several authors studied the PL of disordered materials at room temperature aiming to understand the behavior of elec- tronic levels and the gap variation.4,5,20,21 Periodic quantum mechanical calculations were made within the framework of the DFT using the Becke exchange functional,22 combined with a gradient-corrected correlation functional by Lee, Yang and Parr, B3LYP,23 which has been demonstrated by Muscat et al.24 to be suitable for calculating structural parameters and band structures for a wide variety of solids. CT is a single phase and could be completely indexed on the basis of an orthorhombic ICDD Card No. 42-423 �Pbnm�.25 The unit cell has ten atoms containing two units of CaTiO3 with four nonequivalent atoms in the unit cell having internal experimental values of Ti�0.5,0,0�, Ca�−0.01, 0.03,0.25�, O1�0.07,0.48,0.25�, and O2�0.21,0.21,0.54�. The experimental values for lattice constants a, b, and c are 5.37, 5.44, and 7.64 Å, respectively. As a first step, lattice parameters were optimized to minimize the total energy with respect to the ordered system of the unit cell; Fig. 1�a� shows the CT-o. The calculated lattice constants were a=5.39 Å, b=5.44 Å, and c =7.65 Å for the ordered CT. Our theoretical lattice constant was only �0.004% smaller than the experimental values and is in good agreement with the experimental results. Based on the experimental XANES results presented in this paper, three theoretical models �Fig. 1� were developed to simulate the disordered types and structural complex va- cancies associated with them: �i� displacement of the lattice former, Ti �CT-f�, �ii� displacement of the lattice modifier, Ca �CT-m�, and �iii� simultaneous displacement in the lattice former and modifier, Ti/Ca �CT-fm�. To represent the complex vacancies of oxygen associated with a random lattice of local units �e.g., the TiO6 octahedral complex and CaO12 dodecahedral complex�, it was shifted from its previous position in the unit cell for the lattice former �CT-f model� and lattice modifier �CT-m model�. This displacement causes asymmetries in the unit cell; Ti is now surrounded by five oxygens in a square-based pyramidal configuration, while one other unit cell of Ti is surrounded by six oxygens as in the case of CT-o. Therefore, this asym- metric CT model �CT-f� represents the disorder in the lattice former material. This structure can be designated as �TiO6�– �TiO5·V0 z�, where V0 z =V0 x, V0 • , and V0 ••, which give rise to different complex clusters, depending on the displace- ment of the titanium �Fig. 1�b��. The lattice modifier disorder �CT-m� was modeled by shifting the Ca from its previous position in the modifier unit cell. This displacement causes asymmetries in the unit cell where Ca is now surrounded by 11 oxygens �CaO11·V0 z�, while Ca is surrounded by 12 oxygens �CaO12� as in the case of CT-o. Therefore, this asymmetric CT model �CT-m� rep- resents the disorder in the lattice modifier material. This structure can be designated as �CaO12�– �CaO11·V0 z�, depend- ing on the displacement of the calcium �Fig. 1�c��. Finally, the lattice former and modifier disorder �CT-fm� was modeled at the same time by shifting both titanium and calcium in the same way as described above. Upon heating, three processes leading to crystallization take place in the disordered types and structural complex vacancies �cluster complex� of CT: �a� detachment of �CaO11·V0 z� or �TiO5·V0 z� clusters from underbonded oxygen from their octahedral ��TiO5·V0 z� cluster�; �a� and �b� presumably can occur simul- taneously; �c� redistributing �CaO11·V0 z� clusters and recom- posing all bonds with underbonded oxygen including those produced in �a�. All ab initio calculations were carried out with the CRYS- FIG. 1. �Color online� Structural models for the CT-Pbnm structure. �a� CT-o: ordered and disordered models. �b� CT-f: displacement of the Ti atom. �c� CT-m: displacement of the Ca atom. �d� CT-fm: simultaneous displacement of both atoms. 043526-2 Milanez et al. J. Appl. Phys. 106, 043526 �2009� Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions TAL98 computational code.26 Basis set Ca, Ti, and O centers have been described in the schemes 86-511d3G, 86411-d�31�, and 6-31G�,27 respectively. The Nelder–Mead algorithm28 was used to yield theoretical lattice parameters. To simulate the displacement of the Ti and Ca atoms in the z-direction, the ATOMDISP option was used provided with the CRYSTAL computational code was used. For analysis of the models the results of Mulliken charges,29 electronic levels, optical band gap, and density of states �DOS� were used. IV. RESULTS AND DISCUSSION Ti K-edge XANES spectra of CT samples as a function of the heat treatment temperature �400–600 °C� are shown in Fig. 2�a�. The physical origin of the pre-edge feature A is the transition of the metallic 1s electron to an unfilled d state.30 The A peak area was attributed to the �TiO5� concen- tration and therefore to PL emission.16 Fully structurally or- dered CT does not exhibit PL emission when all Ti atoms are coordinated to six oxygen atoms in a completely regular oc- tahedron �TiO6�. The local order-disorder in the calcium site can be ob- served in Fig. 2�b� by a Ca K-edge. Based on XANES spectra,16 the local disorder in the samples Ca K-edge is re- lated to a change from the 11-fold �CaO11� to the 12-fold �CaO12� coordination number of Ca. These results of the Ca K-edge XANES can be related to the PL emission and the order-disorder in the calcium site, indicating that the modi- fier lattice can strongly affect the intensity of the PL emis- sion due to charge exchanges between �CaO11� and �CaO12� complex clusters. XANES results identified the presence of several com- plex clusters in the disordered CT: �TiO6�, �TiO5�, �CaO11�, and �CaO12�. The existence of such complex clusters con- firms the occurrence of structural defects in the material. Oxygen vacancies are important structural defects that affect the optical property of the solid. In titanates, these vacancies can occur in three different charge states: �1� the �TiO5·V0 x� complex states have captured electrons and are neutral rela- tive to the lattice; �2� the singly ionized �TiO5·V0 • � complex state; and �3� the �TiO5·V0 ••� complex state, which did not trap any electrons and is doubly positively charged with re- spect to the lattice. Another portion of electrons and holes may be trapped by intrinsic crystal defects in the lattice modifier complex �CaO11·V0 z� complex, where V0 z =V0 x, V0 • , and V0 ••. It was speculated that these oxygen vacancies induce new energy levels in the band gap and can be attributed to the calcium-oxygen or titanium-oxygen complex vacancy centers. Direct observation of oxygen vacancies is difficult, but spectroscopic studies of doped and annealed crystals indicate the presence of these isolated vacancies.31,32 Mestric et al.33 considered the presence of six positions for an oxygen va- cancy in ZrO6. It could not be proven experimentally that only six distinct centers are involved. However, by consider- ing the crystal structure, this observation seems to be an obvious assumption. Studies of visible luminescence in TiO2 nanotubes and EPR spectrum show a strong signal at g =2.0034,34 which is characteristic of single-electron-trapped oxygen vacancies.35 CT annealed at 400 and 500 °C shows EPR signals, confirming the presence of �TiO5·V0 • � and/or �CaO11·V0 • � species �Fig. 3�. On the other hand, CT annealed at 600 °C FIG. 2. �Color online� XANES spectra for CT compounds annealed at 400, 500, and 600 °C of �a� Ti K-edge and �b� Ca K-edge. FIG. 3. �Color online� EPR spectra for CT samples annealed at 400, 500, and 600 °C. 043526-3 Milanez et al. J. Appl. Phys. 106, 043526 �2009� Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions does not show EPR signals, confirming the absence of �TiO5·V0 • � and/or �CaO11·V0 • � species. In particular, CT powder annealed at 400 °C presents a major intense feature and an asymmetric EPR line. The spin Hamiltonian parameters of the main contribution were esti- mated by comparison of the experimental spectrum with those generated by a computer simulation program, SINFONIA ®. The calculated values of g� =2.0051 and g� =2.0021 are in agreement with an axial species and are at- tributed to �TiO5·V0 • � and/or �CaO11·V0 • � complex clusters in structurally disorder powders. The intensity of this line was reduced in samples annealed at 500 °C, and it vanished at 600 °C. The presence of different complex clusters is responsible for introducing delocalized electronic levels inside the band gap acting as electron-hole pairs. These electronic levels can be confirmed by UV-vis measurements �Fig. 4�, showing the spectral dependence of the absorbance for the CT samples as a function of the heat treatment temperature. Ordered mate- rials �CT annealed at 600 °C� do not present the intermedi- ary states inside the band gap and display an interband tran- sition. For disordered material �CT annealed at 400 and 500 °C�, this transition does not occur, but absorption edge and tails can be observed, which characterize electronic lev- els inside the band gap.36–38 The optical band gaps obtained according to the Wood–Tauc method39 �Table I� are consis- tent with the interpretation that the exponential optical ab- sorption edge is controlled by the degree of structural order-disorder.19 Figure 5�a� shows PL spectra at room temperature, which were obtained for CT powder annealing at 400, 500, and 600 °C after excitation with a 350.7 nm laser light. The curves display a broadband distributed in the red, green, and blue regions of the visible spectrum �360–800 nm�. In gen- eral, the line shape is typical of the multiphonon process, i.e., it has several relaxation channels, which are indicative of delocalized levels in the band gap. For these disordered ma- terials, the details of band structures in solids are mainly determined by the potential within the unit cell rather than by long range periodicity. Moreover, this line shape indicates that the confinement effect cannot be considered as the pre- dominant mechanism of the luminescence.40 The rearrange- ment of the lattice was clearly detected through PL experi- mental measurements and is a strong indication that this measurement is highly sensitive to structural changes. The PL spectrum for CT annealed at 500 °C is distin- guished by a blue emission. Using the Gaussian method, PL spectra of the CT sample annealed at 400 and 500 °C were deconvoluted into three components: �1� the blue component �maximum below 430 nm�; �2� the green component �maxi- mum below 600 nm�; and �3� the red component �maximum below 730 nm�, corresponding to the regions where the maximum intensity for each component appeared. These de- convolutions represent different types of electronic transi- tions and are linked to a specific structural arrangement. Such electronic transitions are due to the existence of elec- tronic levels in the band gap of a material, which are possi- bly due to structural disorder. Figures 5�b� and 5�c� show results obtained with such deconvolution. These graphs show that with an increase in the thermal treatment time, the red component of PL decreases and the blue component of PL increases. In the annealing process, the disorder powders be- come more ordered with increased annealing times; this structural change also alters the electronic levels in the band gap. The increase in blue emission is important and is attrib- uted to delocalized electronic levels nearest the valence band �VB� and conduction band �CB�.41 With the crystallization FIG. 4. �Color online� Spectral dependence of the absorbance for the CT powder samples annealed at 400, 500, and 600 °C. TABLE I. Results of experimental and theoretical band gaps for CT. Temperature Band gap experimental �eV� 600 °C 3.56 500 °C 3.45 400 °C 2.60 Model Band gap theoretical �eV� Ordered 4.13 Ti 0.5z 3.55 Ca 0.5z 3.29 Ti and Ca 0.5z 3.09 FIG. 5. �Color online� Room temperature PL spectra after exciting CT pow- der with a 350.7 nm laser light. CT annealed at 400, 500, and 600 °C. 043526-4 Milanez et al. J. Appl. Phys. 106, 043526 �2009� Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions process at 600 °C, the defects vanish as well as the PL emis- sion. The maximum intensity of the PL emission is observed for CT heat treated at 500 °C, whereas CT heat treated at 600 °C does not show PL emission, confirming that the PL intensity is associated with the thermal treatment history, structural order-disorder, and electronic levels in the band gap.11,16 PL results can be investigated in terms of the results of the Mulliken charge and cluster vacancies �Table II�. Table II shows that the Mulliken charges for �TiO6� and �CaO12� complex clusters have more negative charges than �TiO5� and �CaO11� clusters. This quantity difference yields a charge transfer among disordered clusters to ordered clusters.36 Therefore, the proposed displacement yields a charge trans- fer from the center to the surrounding area, which is similar to the order-disorder concept,42 i.e., an effect in a localized point of the structure affects neighboring points of this struc- ture. Figure 6 illustrates the calculated values of VB and CB energy levels for ordered and disordered CT models and the difference between �Egap and �Ecx. �Evx shows the behavior of the disordered levels �displaced model� in relation to or- dered levels �crystalline model� for the formation of the in- termediate levels in the decay process of PL emission. In particular, this result is important because it demonstrates the formation of electronic levels with the disorder in the solid and the simultaneous presence of ordered and disordered clusters yielding crystalline defects. The behavior of the VB and CB electronic levels directly alternates the optical band gap of the CT material. Conse- quently, this alteration results in the structural order-disorder. Studies on intermediate levels5 and the order-disorder43 showed a decrease in the optical band gap according to the structural organization. For the decrease in the optical band gap, three forms of displacement of the VB and CB can be proposed: �1� an increase in the VB energy and a decrease in the CB energy; �2� the VB energy remains constant and the CB energy decreases; and �3� an increase in the VB and CB energy remains constant. An increase in the VB energy is shown in Fig. 6, whereas the CB energy can be increased or decreased according to the displacement in the z-direction, demonstrating that the order-disorder has favorable direc- tions for optical properties. Thus, it is possible to theoretically understand the influ- ence of local disorder in the optical band gap in this oxide complex. Therefore, it can be observed in Table I that there are variations between the band gap and order-disorder due to displacements in the molecular structure of the CT mate- rial. These variations are bigger in the band gap as the dis- placement is localized in the z-direction, mainly Ti �z� and Ca �z�. Optical properties with disorder in these directions are more interesting as they favor intermediate levels for a decay process in PL phenomena applicable to optical de- vices. The behavior of these electronic levels can be ob- served by projected DOS atomic calculated results �Fig. 7� for ordered and disordered models with Ti �z� and Ca �z�. There is a main influence in the projected DOS of the O atoms in the VB, whereas the projected DOS of the cations does not have significant variations. These results show that intermediate levels or intermediate states cause a gap de- crease due to the order-disorder effect of the Ca and Ti atoms according to the displacement direction. These studies have shown that order-disordered struc- tures in CT powders obtained by using a soft chemical syn- thesis are two types of coordination for titanium or calcium atoms. Oxygen vacancies in a disordered structure with �TiO6�� / �TiO5·V0 z� and �CaO12�� / �CaO11·V0 z� complex clus- ters are electron-trapping or hole-trapping centers, according to the following equations: �TiO6�x + �TiO5V0 x� → �TiO6�� + �TiO5 · V0 • � , �1� �TiO6�x + �TiO5 · V0 • � → �TiO6�� + �TiO5 · V0 ••� , �2� �CaO12�x + �CaO11 · V0 x� → �CaO12�� + �CaO11 · V0 • � , �3� �CaO12�x + �CaO11 · V0 • � → �CaO12�� + �CaO11 · V0 ••� , �4� where �TiO6�� or �CaO12�� are donors, �TiO5·V0 • � or �CaO11·V0 • � are donors/acceptors �TiO5·V0 ••�, and �CaO11·V0 ••� are acceptors. It is assumed that charge redistri- bution may lead to electron-hole recombination of localized excitons.44 This series of equations represents the complex TABLE II. Results of Mulliken charges �q� for TiO6, TiO5, CaO12, and CaO11 clusters. q�TiO6�� q�TiO5·V0 • � �q1 q�CaO12�� q�CaO11·V0 • � �q2 Ca 0.0 and Ti 0.0 �1.55 �1.55 0 �2.18 �2.18 0 Ca 0.5z and Ti 0.0 �1.56 �1.56 0 �2.19 �1.60 �0.59 Ti 0.5z and Ca 0.0 �1.48 �0.91 �0.57 �2.10 �2.10 0 Ti 0.5z and Ca 0.5z �1.51 �0.94 �0.57 �2.12 �1.50 �0.62 FIG. 6. Illustration of the behavior of the VB and the CB of several atomic displacements in relation to the ordered model. 043526-5 Milanez et al. J. Appl. Phys. 106, 043526 �2009� Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions clusters in structural disordered solids and illustrates the oxy- gen vacancy occurrences that allow interaction between in- terclusters. V. CONCLUSIONS The PL phenomenon of the CT at room temperature is directly influenced by the presence of oxygen vacancies that yield structural order-disorder. These oxygen vacancies bonded at Ti and/or Ca induce new electronic states inside the band gap. One hole in the acceptor and one electron in the donor are yielded in the disordered CT powder. Experimental results indicate that intermediate levels are formed in the band gap for structurally disordered CT. These intermediate levels can be provoked by the oxygen vacancies that can occur in three different charge states �V0 z =V0 x, V0 • , and V0 ••�. These vacancies yield �TiO6�– �TiO5� and �CaO12�– �CaO11� complex clusters that are responsible for PL emission at room temperature. With these theoretical results, it is possible to propose that the atomic displacement is a reasonable approximation for modeling vacancies in the solid state. Vacancies cause structural disorder yielding charged gradient among the lat- tice clusters and hence leading to intermediate levels in the band gap, which is essential for PL phenomena. ACKNOWLEDGMENTS The authors gratefully acknowledge the financial support of the Brazilian research financing Institutions FAPESP/ CEPID �Grant No. 98/14324-0�, CNPq, and CAPES. The research was partially performed at LNLS-National Labora- tory of Synchrotron Light, Brazil. 1H. Pinto and A. Stashans, Phys. Rev. B 65, 134304 �2002�. 2E. Heifets, R. I. Eglitis, E. A. Kotomin, J. Maier, and G. Borstel, Phys. Rev. B 64, 235417 �2001�. 3P. Ghosez, J. P. Michenaud, and X. Gonze, Phys. Rev. B 58, 6224 �1998�. 4E. Longo, E. Orhan, F. M. Pontes, C. D. Pinheiro, E. R. Leite, J. A. Varela, P. S. Pizani, T. M. Boschi, F. Lanciotti, A. Beltran, and J. Andres, Phys. Rev. B 69, 125115 �2004�. 5E. Orhan, J. A. Varela, A. Zenatti, M. F. C. Gurgel, F. M. Pontes, E. R. Leite, E. Longo, P. S. Pizani, A. Beltran, and J. Andres, Phys. Rev. B 71, 085113 �2005�. 6C. M. Liu, X. T. Zu, and W. L. Zhou, J. Phys. D: Appl. Phys. 40, 7318 �2007�. 7E. Pinel, P. Boutinaud, G. Bertrand, C. Caperaa, J. Cellier, and R. Mahiou, J. Alloys Compd. 374, 202 �2004�. 8T. Kyomen, R. Sakamoto, N. Sakamoto, S. Kunugi, and M. Itoh, Chem. Mater. 17, 3200 �2005�. 9X. M. Zhang, J. H. Zhang, X. Zhang, L. Chen, S. Z. Lu, and X. J. Wang, J. Lumin. 122-123, 958 �2007�. 10T. Wanjun and C. Donghua, J. Am. Ceram. Soc. 90, 3156 �2007�. 11A. T. de Figueiredo, V. M. Longo, S. de Lazaro, V. R. Mastelaro, F. S. De Vicente, A. C. Hernandes, M. S. Li, J. A. Varela, and E. Longo, J. Lumin. 126, 403 �2007�. 12B. Ravel and E. A. Stern, Physica B 208-209, 316 �1995�. 13B. Ravel, E. A. Stern, R. I. Vedrinskii, and V. Kraizman, Ferroelectrics 206, 407 �1998�. 14F. M. Pontes, E. Longo, E. R. Leite, E. J. H. Lee, J. A. Varela, P. S. Pizani, C. E. M. Campos, F. Lanciotti, V. Mastelaro, and C. D. Pinheiro, Mater. Chem. Phys. 77, 598 �2003�. 15K. Asokan, J. C. Jan, J. W. Chiou, W. F. Pong, M. H. Tsai, Y. K. Chang, Y. Y. Chen, H. H. Hsieh, H. J. Lin, Y. W. Yang, L. J. Lai, and I. N. Lin, J. Solid State Chem. 177, 2639 �2004�. 16S. de Lazaro, J. Milanez, A. T. de Figueiredo, V. M. Longo, V. R. Maste- laro, F. S. De Vicente, A. C. Hernandes, J. A. Varela, and E. Longo, Appl. Phys. Lett. 90, 111904 �2007�. 17J. R. Sambrano, E. Orhan, M. F. C. Gurgel, A. B. Campos, M. S. Goes, C. O. Paiva-Santos, J. A. Varela, and E. Longo, Chem. Phys. Lett. 402, 491 �2005�. 18E. Orhan, M. Anicete-Santos, M. Maurera, F. M. Pontes, C. O. Paiva- Santos, A. G. Souza, J. A. Varela, P. S. Pizani, and E. Longo, Chem. Phys. 312, 1 �2005�. 19F. M. Pontes, C. D. Pinheiro, E. Longo, E. R. Leite, S. R. de Lazaro, J. A. Varela, P. S. Pizani, T. M. Boschi, and F. Lanciotti, Mater. Chem. Phys. 78, 227 �2003�. 20E. Orhan, F. M. Pontes, C. D. Pinheiro, E. Longo, P. S. Pizani, J. A. Varela, E. R. Leite, T. M. Boschi, A. Beltran, and J. Andres, J. Eur. Ceram. Soc. 25, 2337 �2005�. 21S. M. Hosseini, T. Movlarooy, and A. Kompany, Physica B 391, 316 �2007�. 22A. D. Becke, J. Chem. Phys. 98, 5648 �1993�. 23C. T. Lee, W. T. Yang, and R. G. Parr, Phys. Rev. B 37, 785 �1988�. 24J. Muscat, A. Wander, and N. M. Harrison, Chem. Phys. Lett. 342, 397 �2001�. 25S. Qin, A. I. Becerro, F. Seifert, J. Gottsmann, and J. Jiang, J. Mater. Chem. 10, 1609 �2000�. 26V. R. Saunders, R. Dovesi, C. Roetti, M. Causa, N. M. Harrison, and C. M. Zicovich-Wilson, CRISTAL98 User’s Manual, University of Torino, Torino, Italy, 1998. 27M. Towler, CRYSTAL98 Resources Page, 2007. 28J. A. Nelder and R. Mead, Comput. J. 7, 308 �1965�. 29R. S. Mulliken, J. Chem. Phys. 23, 1833 �1955�. 30R. V. Vedrinskii, V. L. Kraizman, A. A. Novakovich, P. V. Demekhin, and S. V. Urazhdin, J. Phys.: Condens. Matter 10, 9561 �1998�. 31R. Merkle and J. Maier, Phys. Chem. Chem. Phys. 5, 2297 �2003�. 32D. J. A. Gainon, J. Appl. Phys. 36, 2325 �1965�. 33H. Mestric, R. A. Eichel, K. P. Dinse, A. Ozarowski, J. van Tol, L. C. FIG. 7. Project DOS atomic levels for ordered and disordered models. �a� Ordered model. �b� Displacement of the Ti atom in the z-direction. �c� Dis- placement of the Ca atom in the z-direction and �d� displacement of the Ti and Ca atoms in the z-direction and the �x ,y ,z�-directions, respectively. 043526-6 Milanez et al. J. Appl. Phys. 106, 043526 �2009� Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions http://dx.doi.org/10.1103/PhysRevB.65.134304 http://dx.doi.org/10.1103/PhysRevB.64.235417 http://dx.doi.org/10.1103/PhysRevB.64.235417 http://dx.doi.org/10.1103/PhysRevB.58.6224 http://dx.doi.org/10.1103/PhysRevB.69.125115 http://dx.doi.org/10.1103/PhysRevB.69.125115 http://dx.doi.org/10.1103/PhysRevB.71.085113 http://dx.doi.org/10.1088/0022-3727/40/23/011 http://dx.doi.org/10.1016/j.jallcom.2003.11.093 http://dx.doi.org/10.1021/cm0403715 http://dx.doi.org/10.1021/cm0403715 http://dx.doi.org/10.1016/j.jlumin.2006.01.337 http://dx.doi.org/10.1111/j.1551-2916.2007.01929.x http://dx.doi.org/10.1016/j.jlumin.2006.08.100 http://dx.doi.org/10.1016/0921-4526(94)00686-P http://dx.doi.org/10.1080/00150199808009173 http://dx.doi.org/10.1016/S0254-0584(02)00112-8 http://dx.doi.org/10.1016/S0254-0584(02)00112-8 http://dx.doi.org/10.1016/j.jssc.2004.04.043 http://dx.doi.org/10.1016/j.jssc.2004.04.043 http://dx.doi.org/10.1063/1.2713359 http://dx.doi.org/10.1063/1.2713359 http://dx.doi.org/10.1016/j.cplett.2004.12.084 http://dx.doi.org/10.1016/j.chemphys.2004.11.013 http://dx.doi.org/10.1016/S0254-0584(02)00230-4 http://dx.doi.org/10.1016/j.jeurceramsoc.2005.03.053 http://dx.doi.org/10.1016/j.jeurceramsoc.2005.03.053 http://dx.doi.org/10.1016/j.physb.2006.10.013 http://dx.doi.org/10.1063/1.464913 http://dx.doi.org/10.1103/PhysRevB.37.785 http://dx.doi.org/10.1016/S0009-2614(01)00616-9 http://dx.doi.org/10.1039/b000623h http://dx.doi.org/10.1039/b000623h http://dx.doi.org/10.1063/1.1740588 http://dx.doi.org/10.1088/0953-8984/10/42/021 http://dx.doi.org/10.1039/b300205p http://dx.doi.org/10.1063/1.1714477 Brunel, H. Kungl, M. J. Hoffmann, K. A. Schonau, M. Knapp, and H. Fuess, Phys. Rev. B 73, 184105 �2006�. 34L. Qian, Z. S. Jin, J. W. Zhang, Y. B. Huang, Z. J. Zhang, and Z. L. Du, Appl. Phys. A: Mater. Sci. Process. 80, 1801 �2005�. 35E. Serwicka, Colloids Surf. 13, 287 �1985�. 36E. Orhan, F. M. Pontes, M. A. Santos, E. R. Leite, A. Beltran, J. Andres, T. M. Boschi, P. S. Pizani, J. A. Varela, C. A. Taft, and E. Longo, J. Phys. Chem. B 108, 9221 �2004�. 37E. R. Leite, E. C. Paris, F. M. Pontes, C. A. Paskocimas, E. Longo, F. Sensato, C. D. Pinheiro, J. A. Varela, P. S. Pizani, C. E. M. Campos, and F. Lanciotti, J. Mater. Sci. 38, 1175 �2003�. 38P. S. Pizani, E. R. Leite, F. M. Pontes, E. C. Paris, J. H. Rangel, E. J. H. Lee, E. Longo, P. Delega, and J. A. Varela, Appl. Phys. Lett. 77, 824 �2000�. 39D. L. Wood and J. Tauc, Phys. Rev. B 5, 3144 �1972�. 40W. F. Zhang, Z. Yin, and M. S. Zhang, Appl. Phys. A: Mater. Sci. Process. 70, 93 �2000�. 41N. Asha Bhat, K. S. Sangunni, and K. Rao, J. Non-Cryst. Solids 319, 192 �2003�. 42C. Malibert, B. Dkhil, J. M. Kiat, D. Durand, J. F. Berar, and A. Spa- sojevicdeBire, J. Phys.: Condens. Matter 9, 7485 �1997�. 43R. E. Cohen, J. Phys. Chem. Solids 61, 139 �2000�. 44A. T. de Figueiredo, S. de Lazaro, E. Longo, E. C. Paris, J. A. Varela, M. R. Joya, and P. S. Pizani, Chem. Mater. 18, 2904 �2006�. 043526-7 Milanez et al. J. Appl. Phys. 106, 043526 �2009� Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions http://dx.doi.org/10.1103/PhysRevB.73.184105 http://dx.doi.org/10.1007/s00339-003-2486-9 http://dx.doi.org/10.1016/0166-6622(85)80028-7 http://dx.doi.org/10.1021/jp048877y http://dx.doi.org/10.1021/jp048877y http://dx.doi.org/10.1023/A:1022837132220 http://dx.doi.org/10.1063/1.1306663 http://dx.doi.org/10.1103/PhysRevB.5.3144 http://dx.doi.org/10.1007/s003390050018 http://dx.doi.org/10.1016/S0022-3093(02)01918-X http://dx.doi.org/10.1088/0953-8984/9/35/022 http://dx.doi.org/10.1016/S0022-3697(99)00272-3 http://dx.doi.org/10.1021/cm060386y