Journal of The Electrochemical Society, 147 (5) 1993-1996 (2000)
1993
S0013-4651(00)01-025-9 CCC: $7.00 © The Electrochemical Society, Inc.
Thin-Film BaMgAl10O17:Eu Phosphor Prepared by Spray Pyrolysis Nickolay Golego,a,* S. A. Studenikin, and Michael Cocivera**,z Guelph-Waterloo Centre for Graduate Work in Chemistry, University of Guelph, Guelph, Ontario N1G 2W1, Canada For the first time, polycrystalline thin films of europium-doped barium magnesium aluminate (BAM) have been prepared by spray pyrolysis of an aqueous solution of the corresponding metal nitrates. Stoichiometric BAM films were obtained at temperatures as low as 3508C, in sharp contrast to the commonly used high-temperature powder route. The deposition procedure can be applied to other phosphor materials as well, allowing for efficient one-step thin-film phosphor preparation. The prepared films were comprehensively characterized. Their luminescent properties are discussed. © 2000 The Electrochemical Society. S0013-4651(00)01-025-9. All rights reserved. Manuscript received January 7, 2000.
The recent development of commercial phosphors has resulted in remarkable progress in fluorescent lamps and displays, plasma display panels, field emission displays, cathode-ray tubes.1-3 Rare-earth elements in an inorganic oxide host lattice are typically used.4 An important blue-luminescent material is europium-doped barium magnesium aluminate, BaMgAl10O17 or BAM. Commercial production of BAM phosphors involves solid-state reaction of corresponding oxides, carbonates, and hydroxides under a reducing atmosphere at temperatures as high as 16008C for several hours. Such a high temperature is necessary for diffusion of rare-earth ions into the crystal lattice; however, it imposes high cost and energy requirements on the production process. Also, subsequent milling of the solid sintered product has a disadvantage of breaking particles, which reduces phosphor efficiency. Furthermore, to spread and support this powder on a substrate, additional production steps are required using organic binders and firing. These steps often cause phosphor oxidation, which is detrimental to luminescence efficiency. Alternative techniques that can bypass these additional steps, especially the high-temperature sintering, are actively sought. There have been recent reports on sol-gel preparation,5 combustion synthesis,6 and microwave irradiation synthesis7 of BAM powders. An attractive alternative to the present powder route is thin-film deposition, which allows one-step preparation of a thin-film phosphor on a substrate. Surprisingly, little research has been done in this direction, and no reports have been published on thin-film BAM phosphors. Earlier we successfully employed spray pyrolysis to prepare thin films of several oxide compounds such as KTiOPO4,8 RbTiOPO4,9 BaTiO3,10 ZnO,11 and TiO2.12 The present paper reports for the first time the use of spray pyrolysis to prepare thin film BAM doped with Eu. High-temperature rare-earth diffusion is not necessary with this method owing to intimate atomic-scale mixing of the elements in the precursor solution. Effects of post-deposition treatment and luminescent properties of the films are discussed. Experimental The spray-pyrolytic deposition system was described in detail earlier.9,13 Different substrates were used for deposition: Corning 7059 glass, quartz (Chemglass, Inc.), indium-tin oxide (ITO) coated glass (Nesatron, 20 V/u), single-crystalline silicon (Atomergic Chemetals Co.), and alumosilicate ceramic plates (Zircar Products, Inc., type RSDR). Deposition temperature was varied between 120 and 5508C. Humid air was employed as a carrier gas to prevent spray solution evaporation. The stock metal precursor solutions were prepared by dissolving corresponding metal nitrates or oxides in deionized water purified by reverse osmosis. Specifically, Ba(NO3)2, Al(NO3)3, MgO (all 99.999% purity), and Eu2O3 (99.99%) were used to prepare solutions of 0.1 M concentration, which was confirmed by inductively coupled plasma atomic adsorption spectros** Electrochemical Society Student Member. ** Electrochemical Society Active Member. * a Deceased, September 1999. *z E-mail:
[email protected]
copy. The stock solutions were then mixed in an appropriate ratio to ensure a stoichiometric composition of BaMgAl10O17 doped with 10 atom % Eu. Spray solutions of 5-20 mM final concentration were used at an average spray rate of 1 mL per hour. Some films were annealed in air, hydrogen, or forming gas (H2:N2 5 5:95) for 12 h. Morphology and film composition were studied by scanning electron microscopy (SEM) and energy-dispersive spectroscopy (EDS) on a Hitachi S570 scanning electron microscope equipped with a Voyager II 1100 X-ray microanalysis system. Composition data were collected from several spots across the film to confirm that the composition was uniform across the surface within the experimental error of the EDS analysis. X-ray powder diffraction (XRPD) was measured on Rigaku Geigerflex II powder diffractometer with Co Ka radiation. The JCPDS database and PowderCell diffraction pattern simulation software14 were used for phase identification and analysis. UV-vis absorption spectroscopy was done on a Shimadzu UV 160U double-beam recording spectrophotometer at 8 nm/s scan rate. Film thickness and roughness were measured on a Sloan DEKTAK stylus-type recording profilometer as well as from the SEM cross section images. Photoluminescence measurements employed a UV mercury lamp at 254 and 365 nm. The intensity of the lamp was measured with a YSI-Kettering model 65A radiometer to be 0.5 and 2.5 W/m2, respectively. Luminescence spectra were recorded using Ocean Optics S2000 charge coupled device (CCD) spectrometer with a 25 mm input slit and 400 mm single-strand fiber optic feed. Excitation light reflected from the respective uncoated substrates was used as a dark reference. Results and Discussion XRD patterns of BAM films prepared under various conditions are shown in Fig. 1. It is seen that films deposited at and below 3008C (Fig. 1a) are crystalline; however, their diffraction pattern could not be identified as that of BAM. (Note that a broad background peak at 258C is due to the quartz substrate.) Moreover, this pattern completely disappeared upon annealing at higher temperatures as illustrated for films annealed at 800 and 10008C (Fig. 1b and c, respectively). Since BAM is thermodynamically stable in this temperature range, it could not have decomposed during annealing. Therefore, we can conclude that as-deposited films belong to an intermediate, likely a not fully pyrolyzed product such as a mixture of oxides, nitrates, and oxonitrates. Incomplete pyrolysis has been observed for other materials.11,12 This conclusion is supported by the fact that several films deposited at low-temperature partially peeled off the substrate when exposed to humid air for several weeks. Consequently, deposition temperatures at and below 3008C do not provide good quality films. Films deposited above 3008C adhered well and were not affected by moisture or water. They were amorphous by XRD, as evident in Fig. 1d for a film deposited at 4508C. Annealing below 10008C did not induce noticeable crystallization indicating that growth of the crystallites in the material was very slow at these temperatures. This is not unexpected considering the commercial process involved an-
1994
Journal of The Electrochemical Society, 147 (5) 1993-1996 (2000) S0013-4651(00)01-025-9 CCC: $7.00 © The Electrochemical Society, Inc.
Figure 3. SEM images (100 3 100 mm) of thin-film BAM deposited on quartz: (a) as-deposited at 3008C; (b) same film annealed at 10008C for 1 h in air.
Figure 1. XRD patterns of thin-film BAM: (a) as-deposited on quartz at 2008C; (b) annealed at 8008C; (c) annealed at 10008C; (d) as-deposited on glass at 4508C
nealing at temperatures of 16008C. Crystallization was evident in films annealed at 12008C (Fig. 2a). The corresponding as-deposited film, which is shown in Fig. 2b, exhibited only the substrate peaks. It is important to note that the diffraction pattern of the crystallized films (Fig. 2a) closely matched that of BAM (Fig. 2c), therefore, material identity was confirmed. The EDS elemental analysis data also supported the expected stoichiometry of the films. Deposition temperature is known to be a major parameter of the spray pyrolysis technique that influences film adhesion, structure,
Figure 2. XRD patterns of thin-film BAM on alumosilicate ceramics: (a) annealed at 12008C; (b) as-deposited at 5008C, only the substrate peaks are seen; (c) reference BAM, JCPDS card 26-0163.
and porosity.11,12 It was, therefore, of particular importance to establish the deposition temperature regimes for good-quality phosphor preparation. As discussed above, a critical temperature of 3508C that is necessary for full precursor pyrolysis can be inferred from the XRD data. SEM analysis provided support for this conclusion. Films deposited below 3508C often had hairline cracks as illustrated in Fig. 3a for a film deposited at 3008C. Annealing caused film shrinkage and aggravated cracking (Fig. 3b), likely due to material loss during annealing. No such cracking was observed after annealing films deposited at the critical temperature and higher, supporting the conclusion that these films were already fully pyrolyzed. The deposition temperature also affected the film morphology. Figure 4 illustrates the SEM pictures of a series of films grown at various temperatures with the other parameters unchanged. At temperatures below 3008C, thick, sometimes cracked films were grown (Fig. 4a and b). At 3008C films became smooth (Fig. 4c), and at 3508C surface particles started to appear (Fig. 4d). Dense, granular films were obtained in the temperature region from 400 to 5508C (Fig. 4e and f). Cross-sectional SEM analysis showed that films deposited below the critical temperature looked dense but had poor adhesion and integrity (Fig. 5a, note film peeling). Films grown above the critical temperature of 3508C were dense and granular (Fig. 5b) with grain size around several micrometers. Film thickness could be varied between 1 and 10 mm by increasing the deposition time. Films deposited on the granular ceramic substrates acquired a similar granular morphology (Fig. 6a) which was not affected by annealing at 12008C (Fig. 6b). This temperature effect on film morphology also occurred for thin-film titanium dioxide12 although the critical temperature was lower. Optical absorption experiments were complicated by intense light scattering, and only qualitative data could be obtained. Generally, films deposited at and below 3008C did not absorb light in the 3001100 nm region with only slight absorption between 200 and 300 nm. The absorption at shorter wavelength was somewhat improved for films deposited at 3508C and above, as well as for films annealed at 600-10008C. Films annealed at 12008C could not be studied because of the opaque ceramic substrate, although the photoluminescence experiments indicated an increase in UV absorption. It should be noted that good absorption in the UV region is necessary for efficient photoluminescence excitation. The photoluminescence spectra are presented in Fig. 7. Films deposited below 3508C did not luminesce at all, which was in line with the structural data presented above. A different behavior was observed for BAM films deposited at and above 3508C, as well as for all annealed films. The films were not luminescent under 365 nm excitation, in agreement with the optical absorbance data discussed above. However, red luminescence was seen under 254 nm excitation (Fig. 7a). The sharp lines of this luminescence are characteristic for the europium(III) f-f transitions which typically peak around 600-610 nm as, for instance, in Y2O3:Eu3 or oxidized BaMgAl10O17:Eu. Indeed, given the oxidizing conditions involved in film preparation, one expects europium to be in the trivalent state.
Journal of The Electrochemical Society, 147 (5) 1993-1996 (2000)
1995
S0013-4651(00)01-025-9 CCC: $7.00 © The Electrochemical Society, Inc.
Figure 6. SEM images (20 3 20 mm) of thin-film BAM deposited on alumosilicate ceramics: (a) as-deposited at 5008C; (b) annealed in forming gas at 12008C.
crystal lattice which puts the divalent europium ions in a high crystal field, thus increasing the energy difference between the ground and excited levels. This is in agreement with the present work where blue luminescence appeared after annealing in reducing media, which also induced BAM crystallite growth. The disappearance of the red luminescence band indicates that all europium was reduced to the divalent state. This is also supported by the enhanced longwave UV excitation of the blue luminescence which is characteristic for the Eu21 ions. It seems likely that the blue PL is not due to reaction of BAM with the substrate. This substrate consists only of Si, Al, and oxygen. Since BAM is predominantly an aluminate, the aluminum component of the ceramic at most can only decrease the content of Ba, Mg, and Eu in the BAM . Silicon reacts17 with hydrogen between 1200 and 16508C SiO2 1 H2 } SiO 1 H2O
Figure 4. SEM images (20 3 20 mm) of thin-film BAM deposited on glass at different temperatures: (a) 230, (b) 270, (c) 300, (d) 350, (e) 450, and (f) 5508C.
However, since forming gas contains only 5% H2 and our annealing temperature was 12008C, this equilibrium is expected to favor SiO2. Also H2 must penetrate through the BAM without reaction before reaching the substrate. Consequently, the amount of SiO produced is expected to be very small. Furthermore, this species cannot be responsible of the blue PL because the back side of the substrate, which was not covered with BAM and was directly exposed to H2, did not exhibit a measurable PL spectrum. The only other possibili-
After annealing in forming gas at 12008C, the material ceased to luminesce in the red. Instead, an intense blue band appeared under both 254 and 365 nm UV excitation as evident in Fig. 7b and c, respectively. This fact is of practical value because it allows one to use the whole UV spectrum of the mercury discharge instead of only the shortwave part of it. The onset of blue luminescence is thought to be linked to reduction of europium(III) to europium(II) and formation of the BAM
Figure 5. SEM cross-sectional images (10 3 10 mm) of thin-film BAM deposited on quartz at different temperatures: (a) 2008C, arrow indicates substrate surface, and (b) 5508C.
Figure 7. Photoluminescence spectra of thin-film BAM: (a) annealed in forming gas at 12008C, lexc 5 365 nm; (b) same, lexc 5 254 nm; (c) asdeposited at 5008C, lexc 5 254 nm, offset for clarity.
1996
Journal of The Electrochemical Society, 147 (5) 1993-1996 (2000) S0013-4651(00)01-025-9 CCC: $7.00 © The Electrochemical Society, Inc.
that the boundary between the film and the substrate remains well defined and gives no indication of any appreciable interdiffusion between the layers. Under intermediate annealing conditions, orange, yellow, and green hues of luminescence were observed. For instance, annealing in hydrogen at 8008C caused films to luminesce as yellow to the human eye, while same annealing at 10008C induced a shift to “white” luminescence. In our opinion, these are caused by a gradual shift from red to blue luminescence with progressively higher annealing temperatures, although the present data are not sufficient for a quantitative explanation. White luminescence may be attractive for single-material white phosphor applications, and further experiments are planned in this direction. Summary Spray-pyrolytic deposition of thin-film BaMgAl10O17 from an aqueous nitrate solution has been pioneered and reported. Deposition conditions were studied and optimized for efficient blue phosphor preparation. A critical temperature of 3508C, much lower than that used in the powder method, was necessary to prepare films that were identified and characterized as stoichiometric BAM. Films produced below the critical temperature did not have the desired quality and composition. Red luminescence was observed in films deposited at and above the critical temperature, and films annealed at 12008C in forming gas exhibited a bright blue luminescence that could be observed visually in a well-lit room. It should be mentioned that the described method is applicable to other oxide phosphors, including the commercially used green-emitting CeMgAl11O19:Tb and red-emitting Y2O3:Eu, as well as their mixtures with BAM. The reported approach has a great potential for onestep thin-film tricolor phosphor preparation. Acknowledgments This work was supported in part by a grant to M.C. from the Natural Sciences and Engineering Research Council of Canada. The authors wish to thank T. Kolodyazhnyi (Department of Materials Science and Engineering, McMaster University) for the high-temperature annealing. University of Guelph assisted in meeting the publication costs of this article.
References
Figure 8. SEM cross-sectional images of BAM on the alumosilicate substrate: (a) as deposited (200 3 200 mm); (b) annealed in forming gas at 12008C (100 3 100 mm).
ty is diffusion of a small amount of SiO or SiO2 into the BAM film to interact with Eu21 to give the characteristic blue emission of this ion. However, SEM cross section pictures (Fig. 8) of the as-deposited film and of the same film annealed at 12008C in forming gas indicate that interdiffusion is not important. Although the cross section of the ceramic substrate is not smooth, the interface between the substrate and the film is clearly discerned. These pictures clearly show
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.
T. Welker, J. Lumin., 48-49, 49 (1991). C. R. Ronda, J. Lumin., 72-74, 49 (1997). C. R. Ronda, T. Jüstel, and H. Nikol, J. Alloys Comp., 275-277, 669 (1998). B. M. J. Smets, Mater. Chem. Phys., 16, 283 (1987). D. Ravichandran, R. Roy, W. B. White, and S. Erdei, J. Mater. Res., 12, 819 (1997). S. Ekambaram and K. C. Patil, J. Alloys Comp., 248, 7 (1997). D. Dai, Y. Li, S. Cai, Y. Xie, and H. He, J. Rare Earths, 16, 215 (1998). N. Golego and M. Cocivera, J. Electrochem. Soc., 144, 736 (1997). N. Golego and M. Cocivera, Thin Solid Films, 322, 14 (1998). N. Golego, S. A. Studenikin, and M. Cocivera, Chem. Mater., 10, 2000 (1998). S. A. Studenikin, N. Golego, and M. Cocivera, J. Appl. Phys., 83, 2104 (1998). N. Golego, S. A. Studenikin, and M. Cocivera, J. Mater. Res., 14, 698 (1999). N. Golego, Ph.D. Thesis, University of Guelph, Canada (1998). W. Kraus and G. Nolze, Powder Cell 1.8, (1996): http://www.bamberlin.de/a_v/ v_1/powder/e_cell.html 15. S. Oshio, T. Matsuoka, S. Tanaka, and H. Kobayashi, J. Electrochem. Soc., 145, 3903 (1998). 16. E. Luria and S. R. Rotman, J. Lumin., 60-61, 67 (1994). 17. N. C. Tombs and A. J. E. Welch, J. Iron Steel Inst. (London), 172, 69 (1952).
