LUMINESCENCE OF CALCIUM TUNGSTATE PHOSPHORS DOPED WITH EUROPIUM AND TERBIUM

Transcription

1 Moldavian Journal of the Physical Sciences, Vol., N, LUMINESCENCE O CALCIUM TUNGSTATE PHOSPHORS DOPED WITH EUROPIUM AND TERBIUM M. V. Nazarov (), M. V. Zamoryanskaya (), E-J. Popovici (), L. Ungur () B.S. Tsukerblat () () Technical University of Moldova, Kishinev, Moldova () Ioffe Physical -Technical Institute, St-Petersburg, Russia () Raluca Ripan Institute of Chemistry, Cluj-Napoca, Romania () Ben-Gurion University of the Negev, Beer-Sheva, 8 Israel ABSTRACT The cathodoluminescence (CL) modes being incorporated into scanning electron microscope (SEM) give both the spectral and spatial information with high resolution. In this article we apply spectral analysis of CL along with the colour cathodoluminescence (CCL) to the study of calcium tungstate phosphors in a powder form. The samples were prepared by thermal synthesis using the original procedures and were activated by trivalent rare-earth (RE) and ions. Incorporation of + or + -ions in CaWO lattice modifies the luminescence colour due to formation of new luminescence centres. Depending on the nature and concentration of the activator the colour of the luminescence can be varied within the range of visible spectrum so that these materials seem to of promising interest for the applications in optoelectronic devices. We analyze the selection rules for the optical transitions in the RE ion in the site of S point symmetry and the angular dependencies of two-photon absorption. INTRODUCTION The possibility to utilisation the luminescent substances in optoelectronic devices depends on their luminescent properties and sensitivity to various excitation radiations as well as on particle size distribution and crystalline structure of luminous powders. The well-known self-activated calcium tungstate (CaWO :W) is still an efficient material for X-ray image rendering (X-ray intensifying screens) and coloured lighting (advertisements). The characteristics of blue luminescence of these systems are related to the tetrahedral WO groups in the scheelite host lattice. Incorporation of + and/or + -ions into CaWO lattice modifies the luminescence colour due to formation of new luminescence centres. Depending on the activator nature and concentration, the colour of the luminescence is shown to vary within the range of visible spectrum from blue-to-green-to-red [-]. EXPERIMENTAL DETAILS CaWO :RE phosphor samples were prepared by thermal synthesis from homogeneous mixtures consisting of luminescent grade (l.g.) CaWO, l.g. Na WO.H O as flux and l.g. WO and O (99,99% Jansen Chemical) or O (Johnson Mattey Specpure ) as activating system. All luminescent grade substances were prepared in our laboratory by original procedures. Calcium tungstate used as precursor in phosphor synthesis was prepared at room temperature from highly purified CaCl and Na WO solutions. The synthesis of the mixtures containing wt. % flux and equivalent amounts of WO and O or O were wet homogenised, dried at C and fired at 9 C, for hrs; un-covered alumina crucibles were used for the thermal synthesis. The prepared powders were carefully washed, dried and sieved. The incorporation of trivalent rare-earth ions into the CaWO crystalline lattice could proceed in two major ways. Either RE + replace Ca + ions, or a pair RE + + M + replaces Ca + ions (where M + = alkaline ion). Under the conditions adopted in the present work, sodium tungstate acts as mineralising agent and also as a source of compensating ions. As a result, one can assume that during 68

2 Moldavian Journal of the Physical Sciences, Vol., N, the thermal synthesis stage the formation of rare-earth activated phosphor can be described by the following equation: x y x + y x + y ( x y ) CaWO + O + O + WO + Na WO y Ca x y x y Na x + ywo + O 8 Employing this principle we prepared a series ox mixed phosphors with the general formula Ca - x-y x y WO. RESULTS AND DISCUSSION In order to characterize the new prepared samples, the monochromatic and panchromatic CCL- SEM modes were used. It has been shown that the use of colour contrast increases the information content of colour image in comparison with its black-and-white counterpart by about two orders of magnitude []. At the monochromatic mode the CL emission was detected at definite wavelength. At the panchromatic CL-mode the CL was recorded at all the visible emission spectrum. By this way, monochromatic CL corresponds to definite transitions between the defect levels attributed to the activator. Spatial distribution of the monochromatic or panchromatic CL on the surface under study was recorded in forms of corresponding CCL-maps. A variety of colours observed in panchromatic CCLmap show, therefore, a spatial distribution of all the luminescent centres that occurred by nature or design on the surface of investigated phosphors (fig.). ig.. Panchromatic images of CaWO activated with + and + (original red colour for incorporations and green for ) To explain the nature of CCL and to reveal the spectral composition and its variations with RE-activator, the PL and CL spectra were recorded. The PL (ig. ) and CL spectra (ig. ) of the same material are rather similar, however, there is some difference between the PL and CL related with excitation, generation rate and volume. In general, the CL spectra are richer because only the electron beam excitation can induce transition between all the defect levels, including those created by surface type of imperfections. Observed characteristic blue emission at - nm is related to some tetrahedral WO groups of the scheelite host lattice. The specific green emission bands at 9 and nm can be attributed to the electronic transition in the -centre. red emission bands at 96, 6, 66, nm are caused by transition from emitting level D to,,, for trivalent ions of europium (ig. ) 69

3 Moldavian Journal of the Physical Sciences, Vol., N, ig..pl spectra of calcium tungstate phosphors varied with activators and its concentration D - CL Intensity (a.u.) 6 D - D - D - 6 D - D - 6 D - D - D - D - D λ, nm ig.. CL spectra of CaWO, activated with the trivalent ions of + and +

4 Moldavian Journal of the Physical Sciences, Vol., N, In order to clarify the origin of the major changes produced by different incorporated activators on the apparent luminescence colour and brightness the numerous experiments were performed and the - level model taking into account the interaction between and was proposed. E, X - см - D + + D ν= см - D D λ, nm ig.. Schematic representation of the energy transfer in CaWO, activated with the trivalent ions of + and +

5 Moldavian Journal of the Physical Sciences, Vol., N, The luminescence spectra of the systems under consideration are shown in igs. and along with the assignation of the different groups of the lines arising from the transitions D, D J in + + and D J in. The refinement of the structure of CaWO by Zalkin and Templeton [] gave excellent structure data. The space group for the tetragonal crystals is 88 (I a, C 6 h ) with tungsten in a position, having a site symmetry S. (see fig. ). ig.. a) Crystal structure of CaWO (scheelite) and b) CaWO, activated with + and + It is surrounded by four oxygen ions in a nearly regular tetrahedron at distances of.8 nm. Calcium ions have eight oxygen neighbours, at. nm and at.8 nm, forming a distorted cube. Each tungsten ion has four calcium ions above and below this plane at distance of.8 nm. Replacing one or more of these calcium ions by or or other incorporated ions will distort the WO tetrahedron by changing the position of the linking oxygen ion. We can thus explain the shifts of the wavelength of maximum intensity in blue region of spectra (see fig. for example) as a small amount of doped ions replace Ca in WO as being due to increasing distortion of the WO complex. The analysis of the of the optical lines is based on the calculation of the selection rules for the transitions and crystal field pattern. Hereunder we will focus on the former problem and consider also the angular dependencies for the two-photon transitions. + ion contains 6 f-electrons, so the terms formed by the f 6 -shell belong to the even + representations D J ( sign + ). A ( Γ ) and B ( ) point group E ( Γ + ) Γ are two one-dimensional representation of S Γ - two-dimensional representation (we use Malliken s notations, Bete s symbols are given in parentheses, Γ, Γ one-dimensional complex-conjugated representations corresponding to the orbital doublet E). The components of the dipole moment operator d correspond to the following irreducible representations: B (d z ), E (d x, d y ), where z-axis is parallel to C axis in S point group, x and y are perpendicular to C. The component d z gives rise to the light emission (or absorption) in z- polarization ( - polarization), the components d x, d y are related to x and y polarizations, i.e. in the plane that is perpendicular to the main axis C ( -polarization). This assignment will be used in the analysis of the selection rules and polarization of the allowed transitions between Stark components of the free + ion in the crystal field of S symmetry. Our aim is to calculate the selection rules and polarization for all transitions in +

6 Moldavian Journal of the Physical Sciences, Vol., N, D 6,,,,,, each being split accordingly to the set of the irreducible representations of S for a free ion terms as listed in Table. Table Splittings of S+ D J multiplets of + in the crystal field of S symmetry. Low-lying terms + D J Scheme of the reduction of in the point symmetry group S [,6] A + B + E 6 A + B + E A + B + E A + B + E A + B + E A + E A Using the multiplication table one can find out the following rules: ) A A transitions are strictly forbidden, ) A B transitions are allowed and the optical lines are polarized along C axis ( -polarization), ) A E transitions are also allowed and the corresponding lines are polarized perpendicular to C axis ( -polarization). The results are illustrated in ig., where each transition we show schematically fine structure and polarization rules caused by the crystal field splitting for a particular D J optical line. or example, the spectral pattern for the transition D 6 is expected to give rise to lines, of them are -polarized ( A B transitions), lines ( A E ) are -polarized, transitions of A A type are forbidden. It should be emphasized that the above consideration is based on the group-theoretical arguments and so the positions of the lines and their relative intensities are out of the scope of this consideration. In this view it should be noted that the mutual disposition of the Stark levels is only conventional and is aimed to illustrate of the expected number of the lines and their polarization. Nevertheless, combining this consideration with the experimental data on single crystals at different temperatures one can arrive to the important conclusions concerning the assignment of the observed fine structure of the emission spectrum. Unfortunately, presently these experimental data are not available. Regarding the complicated crystal field spectra of rare-earth ions the two-photon laser spectroscopy should be mentioned as an additional power tool in studying of these spectra that was demonstrated in [8, 9] (see [, ] ). irstly, the restriction implied by the selection rules for onephoton transitions would be removed. As an example the forbidden one-photon A A transition can be mentioned. This transition proves to be allowed in the two-photon process if both photons are polarized along C axis, in fact, B( d z ) B( dz ) = A, so that A A transition is allowed. This transition is also allowed if both photons are polarized in the plane. This can be seen from the decomposition of the direct product E( dx,d y ) E( dx,d y ) = A + B that contains A. At the same time this transition can be referred to as the forbidden if one photon is polarized along C axis and the second one in plane ( E( d,d ) B( d ) E does not contain A). This property can serve as a tool for the x y z =

7 Moldavian Journal of the Physical Sciences, Vol., N, assignation of this particular transition. More general and detailed information can be obtained with the aid polarization dependencies of two-photon transitions, i.e. dependencies of the intensities from the components of polarization vectors in two laser beams. In order to find these dependencies for the two-photon transitions in S point group we shall use the general approach developed in [] and exploiting full advantage from the symmetry arguments (see analysis in []): Γ Γ ( u,u ) V V ( u) Γ Γ = Γ Γ γ Γγ I () In eq.() the following notations are used: Γ and Γ are the electronic terms in crystal field (irreducible representations of point symmetry group), Γ Γ Γ, u and u polarization vectors of two photons whose components are transformed like the components of the dipole moment, V Γγ ( u) are the symmetry adapted components of the second rank tensor constructed from the components of ΓΓ vectors u and u ( l = ux, m = u y, n = uz, etc.), γ enumerates the basis of Γ. The V Γ are independent (from the point of view of symmetry) parameters and they are related to the irreducible representations Γ. These parameters can be considered as the semiempirical parameters to determined by the comparison of the calculated angular dependencies with the measured ones. The tensors V Γγ ( u) can be built with the aid of the Clebsch-Gordan decomposition: VΓγ ( u ) = u( Γd ) u( Γd ) Γd γγd γ Γγ () γγ where Γd γγd γ Γγ are the Clebsch-Gordan coefficients for the point group under consideration, Γd - irreducible representations which the dipole moment components belong to. Only those representations Γ should be taken in eq.() that belong to the intersection Γ Γ Γ Γd Γd that is the condition for the two-photon transitions to be allowed. The number of the irreducible representations Γ belonging to the intersection Γ Γ Γ Γd Γd determines the number of the independent semiempiric parameters V Γ Γ Γ in the equation for the angular dependence. The application of the mentioned procedure leads to the following results (they are given only for the transitions involved in the luminescence spectrum of +, ig.6): Transition A A: I AA (,u ) = n n + c ( l l + m m ) + c ( l m m ) Transition A B: u. () l (,u ) = ( l l m m ) + c ( l m m ) I AB u +. () Transition A E: l ( u,u ) = n n + c ( l l + m m ) + c ( l m m l ) + c ( l l m m ) + c ( l m m ) I AE + 6 l. () In eqs.()-() c i are the semiempirical parameters. If the experimental data on the intensities vs. angles of polarization vectors are available these relations enable us to make reliable assignations of the spectral lines to the transitions between the certain Stark components in the spectrum. In particular, from eq. () one can see that A A and A B lines disappear if one photon is polarized along C axis and the second one in the plane. On the contrary, A E lines are observable in all cases, namely, when both photons are polarized along or perpendicular to C axis and when one photon is polarized along

8 Moldavian Journal of the Physical Sciences, Vol., N, C and the second one in the plane. Selecting the polarization vectors along C axis one can observe only A A and A E transitions meanwhile A B lines are expected to disappear. The angular dependence is peculiar to each kind of the transition. A D A B 6 E A B E A B E A B E A B E A E ig. 6. Splitting of S+ D J multiplets of + in the crystal field of S symmetry A

9 Moldavian Journal of the Physical Sciences, Vol., N, The rules describing polarization in two photon transitions are related to the purely electronic lines, meanwhile the optical pattern involves also vibronically assisted transitions. Usually for the f shells of the rare-earth ions that are well screened from the surrounding the vibronic interaction is weak so that the conditions for the observation of the well developed vibronic structure are favourable. That is why in many cases it is difficult to distinguish between the electronic lines and their vibronic satellites in the overall spectrum. In these cases the angular dependencies of the light absorption (in one- and two-photon processes) in the region of the vibronic satellites can serve as an additional tool for the identification of the lines. We are going to discuss this problem elsewhere. or the two-level system (excited state and ground state) the CL intensity from excited level () is determined by the population of this level. The population of the excited state decreases according to dn = LJn τ n (6) dt where n and n gives the number of luminescent ions in the ground () and excited () state, τ - is the probability for spontaneous emission ; J- is an electron beam density, L is a quantum yield of CL. Taking into account that for the stationary conditions of excitation dn/dt =, and the total number of luminescent ions N=n +n, the equation () looks as follows: LJN n = () LJ + τ Obtained expression explains linear dependence of CL intensity vs. activator concentration at low concentrations when the interaction between ions is negligible. At the high ions concentration, and especially when different ions and are presented together, a more complicated three-level model could be proposed. In this case + serves as donor and + ions serve as acceptor and one additional channel for population of acceptor level is appeared. and ions concentration determine the probability of energy transfer. The most efficient energy transfer corresponds to the case when the concentrations of activators are equal. We can assume that probability of this process will be given by N N K = (8) N where N is a total and concentration. The additional energy transfer from the excited level D decreases the life- time of this excited state. Then for the stationary conditions of excitation the following expression will be applied LJn [ τ + ( ) ] / N NN n = (9) where n and n populations of ground state and excited state of ions. Considering that n +n =N, and N +N =N, the expression for the population of the excited can be found as: LJN I ~ n = () LJ + τ + (/ N ) N ( N N ) or ions the energy transfer from means an availability of one additional channel. At that time for the stationary conditions of excitation the next expression will be written: [ ( / N ) N N ] LJn τ n = () where n and n - populations of ground state and excited state of ions. Taking into account that n +n =N, and N +N =N, the equation for excited level will be given by LJN I ~ n = () LJ + τ (/ N ) N ( N N ) Comparing () и (), we can conclude that when an interaction between and ions is considered, the luminescence intensity from centres is higher then from ones. If the condition LJ>>τ - - (/N )N (N-N ) is satisfied, the luminescence intensity can be described as 6

10 Moldavian Journal of the Physical Sciences, Vol., N, I~n =N () Hence, according to proposed model the emission intensity of decreases nonlinearly and increases linearly with activator concentration. To verify this theory, some experiments with powder phosphors CaWO, activated with + and + with different concentrations from to moles % were carried out (fig.). I CL - - % - % % - % 6 - % - % % -.% - % - % λ, nm λ, nm λ, nm λ, nm 6 λ, nm ig..cl spectra of CaWO : and CaWO : with different activator concentrations

11 Moldavian Journal of the Physical Sciences, Vol., N, Concentration dependence of these compounds Ca -x-y x y WO is given in fig. 8, 9. When + and + were introduced separately, the linear dependence may be observed (fig.8). CL Intensity (a.u.) CL Intensity (a.u.) X, at.% X, at.% ig.8. CL intensity vs. activator concentration. and introduced separately. ig.9. CL intensity vs. activator concentration and introduced in equal ratio When + and + were introduced together in equal ratio, at lower concentration (up to -,%) we find also the linear dependence (fig.9) and their interaction cannot be accounted. We can use in this case the simple two-level model (formulas (6,) ). At higher concentration the activator interaction must be considered to explain the results and equations ()-() are applied. CONCLUSION CL-SEM method could be proposed for a complex investigation of phosphorescence and fluorescence from materials with different structures. CL-images obtained in SEM in real colours can be used for a rapid examination of spatial and spectral characteristics of powder materials, produced by different technologies. The proposed model taking into account an interaction between different activators is in a good agreement with experimental data. Group-theoretical consideration is given for the selection rules and the angular dependencies of two photon transition in the + spectrum in the site with S point symmetry. New luminescent system based on the powder CaWO : + and CaWO : + phosphors with various luminescence colours and excitability prepared in the classic ceramic techniques can be employed in the manufacture of the optoelectronic devices. ACKNOWLEDGEMENT B.S. Tsukerblat thanks Supreme Council on Science and Technological Development of Moldova (grant ) for the financial support. AUTHOR DETAILS Dr. M. Nazarov, Department of Physics, Technical University of Moldova, Blvd. Stefan cel Mare 68, Kishinev, MD-, Moldova, Tel: +()---8, mvnazarov@mail.ru REERENCES [] T. A. Nazarova, M. V. Nazarov, G.V. Saparin, S.K. Obyden, I.P Ivannikov, E.J. Popovici,. orgaciu, Colour Cathodoluminescence and Photoluminescence of Powder Samples. /Microscopy and Analysis, March,, P.9-. 8

12 Moldavian Journal of the Physical Sciences, Vol., N, [].orgaciu, E.J.Popovici and M.Vasilescu. Study on the synthesis of rare-earth activated calcium tungstate phosphors. //Proceedings of SIOEL 98, Bucharest 998, P.8-8, [].orgaciu, E.J.Popovici, L.Ungur, M.Vadan, M. Vasilescu and M.Nazarov, " On the synthesis of europium or terbium activated calcium tungstate phosphors, //SPIE Proceedings Series,,, P. -9. [] Saparin G.V. and Obyden S.K. Colour in the Microworld: Real Colour Cathodoluminescence Mode in Scanning Electron Microscopy. /ropean Microscopy and Analysis, March. -9, 99 [] Zalkin A. and Templeton D. J.Chem. Phys., P., 96 [6] G..Koster, J.O.Dimmok, R.G.Wheeler, H.Statz, Properties of the Thirty-Two Point Groups, M.I.T. Press, Cambridge, Massachusets, 96. [] B.S.Tsukerblat, Group Theory in Chemistry and Spectroscopy, Academic Press, London, 99. [8] M.Inoue, Y.Toyazawa, J.Phys.Soc.Japan, 96,, P.6. [9] T.R.Bader, A.Gold, Phys.Rev., 968,. [] B.S.Tsukerblat, E.V.Vitiu, in: Optical and Kinetic Effects in Strong Electromagnetic ields, Shtiintsa, Kishinev, 9, P.8-6 9