Langmuir 2007, 23, 3947-3952
3947
Preparation of Nd:YAG Nanopowder in a Confined Environment Eugenio Caponetti,* Delia Chillura Martino, and Maria Luisa Saladino Dipartimento di Chimica Fisica “F. Accascina”, UniVersita` di Palermo and INSTM UdR Palermo, Viale delle Scienze, Parco d’Orleans II, Pad. 17, I-90128 Palermo, Italy
Cristina Leonelli Dipartimento dei Materiali ed Ingegneria Ambientale, UniVersita` di Modena e Reggio Emilia, 41100 Modena, Italy ReceiVed September 4, 2006. In Final Form: January 15, 2007 Nanopowder of yttrium aluminum garnet (YAG, Y3Al5O12) doped with neodymium ions (Nd:YAG) was prepared in the water/cetyltrimethylammonium bromide/1-butanol/n-heptane system. Aluminum, yttrium, and neodymium nitrates were used as starting materials, and ammonia was used as a precipitating agent. Coprecipitate hydroxide precursors where thermally treated at 900 °C to achieve the garnet phase. The starting system with and without reactants was characterized by means of the small-angle neutron scattering technique. The system, without reactants, is constituted by a bicontinuous structure laying near the borderline with the lamellar phase region. The introduction of nitrates stabilizes the bicontinuous structure, while the presence of ammonia induces a transformation from the bicontinuous phase to a lamellar phase. Nd:YAG nanopowder was characterized by wide-angle X-ray scattering, transmission electron microscopy, gas adsorption, and photoluminescence spectroscopy. By comparison with a sample prepared by the conventional coprecipitation method, the obtained Nd:YAG nanopowder is constituted by smaller crystalline nanoparticles showing a lower tendency to agglomerate. In addition, the nanoparticles present a welldefined spherical shape. Photoluminescence spectroscopy confirms that the doping Nd3+ ions substitute Y3+ ions in the YAG crystalline lattice. The Nd3+ lifetime value, obtained from the luminescence decay curves, was 286 ( 10 µs, higher than the single-crystal value (255 µs) and much higher than the nanopowder value obtained by the conventional coprecipitation method (75 µs).
Introduction Yttrium aluminum garnet (YAG) has the chemical formula Y3Al5O12. Nanosized materials doped with lanthanide ions (Ln: YAG) exhibit properties, in particular luminescence ones,1,2 significantly different from those observed in the bulk. The usual applications of the above materials include phosphors and laseractive media. Eu:YAG and Tb:YAG are red and green wellknown phosphors, respectively.3,4 The codoped Er,Yb:YAG is used as an upconversion phosphor.5 The Nd:YAG single crystal has proved to be an outstanding solid-state laser material,6-9 and it has recently been used to produce highly efficient transparent ceramics.10 Recent investigations indicate that Nd:YAG polycrystalline ceramics based on sintering of nanopowders are one of the most promising materials for solid-state lasers.11-13 * To whom correspondence should be addressed. Phone: (+39) 0916459842. Fax: (+39) 091-590015. E-mail:
[email protected]. (1) Yoffe, A. D. AdV. Phys. 1993, 42,173. Bhargava, R. N. J. Lumin. 1996, 70, 85. (2) Fu, X.; Qutubuddin, S. Colloids Surf., A 2001, 179, 65. (3) Zhang, J. J.; Ning, J. W.; Liu, X. J.; Pan, Y. B.; Huang, L. P. Mater. Lett. 2003, 57, 3077. (4) Zhang, J. J.; Ning, J. W.; Liu, X. J.; Pan, Y. B.; Huang, L. P. J. Mater. Sci. Lett. 2005, 22, 13. (5) Liu, M.; Wang, S. W.; Zhang, J.; An, L. Q.; Chen, L. D. Key Eng. Mater. 2005, 517, 80. (6) Geusic, J. E.; Marcos, H. M.; Van Uitert, L. G. Appl. Phys. Lett. 1964, 4, 182. (7) Kaminskii, A. A. Crystalline Lasers: Physical Processes and Operating Schemes; CRC Press: Boca Raton, FL, 1996; Chapter 1. (8) Kaminskii, A. A.; Vylegzhanin, D. N. IEEE J. Quantum Electron. 1971, 7, 329. (9) Kimura, T.; Otsuka, K.; Saruwatari, M. IEEE J. Quantum Electron. 1971, 7, 225. (10) Ikesue, A.; Kinoshita, T.; Kamataka, K.; Yoshida, K. J. Am. Ceram. Soc. 1995, 78, 1033. (11) de With, G.; van Dijk, H. J. A. Mater. Res. Bull. 1984, 19 (10), 1669. (12) Mudler, C. A.; de With, G. Solid State Ionics 1985, 16, 81.
The synthesis of doped YAG nanopowders is the current subject of several studies. A number of synthetic methods have been suggested with preference for soft, low-temperature chemical routes, because these allow a superior control of the powder purity, homogeneity, and physical properties.14-20 To obtain the YAG nanopowder, the hydroxide precursors must be annealed at high temperature (900-1200 °C). The thermal treatment can induce grain growth and/or cluster aggregation; therefore, it is a determining step for the crystalline structure of the material, but it also influences significantly the nanoparticles’ size and their aggregation. The production technology of high-quality YAG nanopowders, however, is still subject to intense studies since the optical properties of nanocrystals are strongly dependent on the doping amount, size, morphology, and synthesis conditions. Therefore, it is desirable to implement existing methods and to search for new ones to produce nanoparticles having the desired size, morphology, structural characteristics, and physical properties. The synthesis of nanoparticles in confined environments has been widely reported in the literature.21,22 Among the various (13) Sekita, M.; Haneda, H.; Yanagitani, T.; Shirasaki, S. J. Appl. Phys. 1990, 67 (1), 453. (14) Yamaguchi, O.; Matui, K.; Shimizu, K. Ceram. Int. 1985, 11, 107. (15) Gowda, G. J. Mater. Sci. Lett. 1986, 5, 1029. (16) Yamaguchi, O.; Takeoka, K.; Hirota, K.; Takano, H.; Hayashida, A. J. Mater. Sci. 1992, 27, 1261. (17) Wang, H.; Gao, L.; Niihara, K. Mater. Sci. Eng. 2000, A288, 1. (18) Zhang, X.; Liu, H.; He, W.; Wang, J.; Li, X.; Boughton, R. I. J. Alloys Compd. 2004, 372, 300. (19) Liu, J.; Ueda, K.; Yagi, H.; Yanagitani, T.; Akiyama, Y.; Kaminskii, A. J. Alloys Compd. 2002, 341, 220. (20) Palmero, P.; Esnouf, C.; Montanaro, L.; Fantozzi, G. J. Eur. Ceram. Soc. 2005, 25, 1565. (21) Vaucher, S.; Fielden, J., Li, M.; Dujardin, E.; Mann, S. Nano Lett. 2002, 2 (3), 225.
10.1021/la0625906 CCC: $37.00 © 2007 American Chemical Society Published on Web 02/28/2007
3948 Langmuir, Vol. 23, No. 7, 2007
systems, water in oil microemulsions have aroused the interest of researchers.23-27 The term microemulsion, in its most general use, denotes an oil/water mixture thermodynamically stabilized by a third component, a surfactant, able to reduce the water-oil interfacial tension. An alcohol is often necessary for microemulsions to form. Over the past three decades, it has been shown that microemulsions are structurally well-defined self-organized systems capable of assuming a variety of microstructures such as water-in-oil droplets (w/o microemulsions or L2 phases), droplets of oil in water (o/w microemulsions or L1 phases), bicontinuous as well as lamellar phases, and so on. w/o microemulsions have been used for the synthesis of a wide variety of nanoparticles, such as metals, alloys, sulfides, and other kinds of materials.25,28-30 Because of the limited size of the water pools (2-10 nm), they result in control of the size and polydispersity of the nanoparticles, even if the final size of the material unlikely coincides with those of microheterogeneities present in the synthesis environment. By exploiting the precipitation in microemulsion media, some mixed-metal oxides have recently been prepared from specific precursors: in particular, barium ferrite, BaFe12O19, was obtained by coprecipitation of carbonate particles;31 high-temperature superconductors, YBa2Cu3O7-x, and some perovskites, such as LaNiO3, La2CuO4, and BaPbO3, have been synthesized by coprecipitation of oxalate particles.32,33 In all cases the reaction takes place when two microemulsions containing the reactants are mixed. For the above pourpose, the w/o microemulsions more frequently used are cetyltrimethylammonium bromide (CTAB)/ 1-butanol/octane/water, sodium octanoate/decanol/water, polyethylene glycol dodecyl ether (PEGDE)/hexane/dodecane/water, water/isooctane/sodium bis(2-ethylhexyl)sulfosuccinate (AerosolOT, AOT), water/cyclohexane/AOT, and others. In the past few years, our interest was focused on the synthesis of YAG nanopowders doped with lanthanides for technological applications such as the production of transparent ceramics and white light emission diodes (LEDs). Recent investigations suggested that a nanopowder composed by unaggregated spherical particles is the most suitable material for compacting and sintering of ceramics and for obtaining a homogeneous dispersion of nanoparticles in a transparent medium.34,35 Most of our attention has been devoted to the study of the structural and optical properties of Nd:YAG nanopowder.36,37 On the basis of our previous works,36-38 in this paper we report the use of the CTAB/ 1-butanol/n-heptane/water microemulsion to obtain Nd:YAG nanopowder having the desired properties. In addition we present (22) Valentini, L.; Bavastrello, V.; Armentano, I.; D’Angelo, F.; Pennelli, G.; Nicolini, C.; Kenny, J. M. Chem. Phys. Lett. 2004, 392, 214. (23) Petit, C.; Lixon, P.; Pileni, M. P. J. Phys. Chem. 1990, 94, 1598. (24) Wilcoxon, J. P.; Williamson, R. L.; Baughman, R. J. Chem. Phys. 1993, 98, 9933. (25) Capek, I. AdV. Colloid Interface Sci. 2004, 110, 49. (26) D’Aprano, A.; Pinio, F.; Turco Liveri, V. J. Solution Chem. 1991, 20, 301. (27) Fu, X.; Qutubuddin, S. Colloids Surf., A 2001, 179, 65. (28) Nagy, J. B.; Claerbout, A. In Surfactants in Solution; Mittal, K. L., Lindmann, B., Eds.; Plenum Press: New York, 1990; Vol. 11, p 366. (29) Lopez-Quintela, M. A.; Rivas, J. J. Colloid Interface Sci. 1993, 158, 446. (30) Petit, C.; Pileni, M. P. J. Phys. Chem. 1988, 92, 2282. (31) Pillai, V.; Kumar, P.; Shah, D. O. J. Magn. Magn. Mater. 1992, 116, L299. (32) Gan, L. M.; Zhang, L. H.; Chan, H. S. O.; Chew, C. H.; Loo, B. H. J. Mater. Sci. 1996, 31, 1071. (33) Ayyub, P.; Maitra, A. N.; Shah, D. O. Physica C 1990, 168, 571. (34) Fang, C. S.; Chen, Y. W. Mater. Chem. Phys. 2003, 78, 739. (35) Moon, Y. T.; Park, H. K.; Kim, D. K.; Kim, C. H. J. Am. Ceram. Soc. 1995, 78, 2690. (36) Caponetti, E.; Saladino, M. L.; Chillura Martino, D.; Pedone, L.; Enzo, S.; Russu, S.; Bettinelli, M.; Speghini, A. Solid State Phenom. 2005, 106, 7. (37) Caponetti, E.; Enzo, S.; Lasio, B.; Saladino, M. L. Opt. Mater., in press. (38) Caponetti, E.; Pedone, L.; Chillura Martino, D.; Panto`, V.; Turco Liveri, V. Mater. Sci. Eng., C 2003, 23 (4), 531.
Caponetti et al.
Figure 1. SANS experimental data (points) vs Q for (A) a microemulsion without reactants, (B) a microemulsion containing aluminum, yttrium, and neodymium nitrates, and (C) a microemulsion containing ammonium hydroxide. Dotted lines represent the calculated intensities obtained by using the ellipsoidal particles interacting via the hard-sphere potential model (A) and the bicontinuous model (C); continuous lines were obtained by using the bicontinuous model (A, B) and the lamellar phase model (C) described in the text.
and discuss results obtained by a structural investigation performed on both the Nd:YAG nanopowder and its synthesis environment. The CTAB/1-butanol/n-heptane/water microemulsion has been used to produce hydroxide precursors of Nd0.15Y2.85Al5O12 (5% replacement of Y with Nd). This microemulsion, contrarily to the water/AOT/n-heptane39,40 microemulsion used by us for the synthesis of CdS nanoparticles,38 does not suffer hydrolysis as a consequence of hydroxide presence. Hydroxide precursors, successfully obtained, were calcined at 900 °C to get the YAG phase that has been characterized by means of wide-angle X-ray diffraction (WAXS), field-emission gun scanning electron microscopy (FEG-SEM), transmission electron microscopy (TEM), photoluminescence (PL) spectroscopy, and specific surface area measurements. A comparison with the nanopowder obtained by using the coprecipitation method36 is also presented. (39) Fletcher, P.; Perrins, N.; Robinson, B.; Toprakcioglu, C. In ReVerse Micelles; Luisi, P. L., Straub, B. E., Eds.; Plenum Press: New York, 1984; p 69. (40) Delord, P.; Larche, F. J. Colloid Interface Sci. 1984, 98, 277.
Preparation of Nd:YAG Nanopowder
Figure 2. WAXS pattern of Nd:YAG nanopowder calcined at 900 °C for 1 h and 1 + 1 h. YAH phase peaks are indicated by open circles.
The structure of the synthesis environment has been investigated by the small-angle neutron scattering (SANS) technique, performing measurements on the microemulsion without and in the presence of reactants. This is to understand how the presence of reactant alters the synthesis environment and, eventually, how this determines the final properties of the produced material. To our knowledge, in the literature only information on the state of the above microemulsion (i.e., the reversed phase or bicontinuous phase) obtained by means of conductivity measurements41 is reported. Experimental Section Materials. Y(NO3)3‚6H2O (Aldrich, 99.9%), Al(NO3)3‚9H2O (Aldrich, 98%), and Nd2O3 (Sigma-Aldrich, 99.99%) were the sources of Y3+, Al3+, and Nd3+ ions, respectively. Nitric acid (Aldrich, 90%), ammonia solution (E. Merck, 25%), 1-butanol (Aldrich, 99.8.%), n-heptane (Aldrich, 99%), CTAB (Aldrich), and D2O (Aldrich, 99.8% deuterium) were used as received. Solutions were prepared by weight and by adding conductivity grade water or D2O. Nd:YAG Nanopowder Preparation. A neodymium, yttrium, and aluminum nitrate solution (0.15 × 10-2, 2.75 × 10-2, and 5.0 × 10-2 mol L-1, respectively) was prepared by dissolving Nd2O3 in dilute nitric acid and, then, by adding Y(NO3)3‚6H2O and Al(NO3)3‚9H2O. Two CTAB/1-butanol/n-heptane/water microemulsions, A and B, differing only in the aqueous phase, were prepared. The aqueous phase of microemulsion A was the nitrate solution, whereas the aqueous phase of microemulsion B was an ammonium hydroxide 5.0 mol L-1 solution. The composition of the four-component microemulsion is defined by the CTAB concentration (7.44 × 10-1 mol kg-1), the water/surfactant molar ratio, R, and the cosurfactant/ surfactant molar ratio, P (R ) 57.3 and P ) 4.1). Microemulsion B was added dropwise to microemulsion A maintained under constant stirring at room temperature until an apparent pH of 8 was reached. A white hue sol was instantaneously observed, indicating the formation of hydroxides. A complete precipitation occurred in 12 h. The mixed neodymium-yttriumaluminum hydroxide precipitate was filtered and repeatedly washed with water to remove residual ammonia, nitrate ions, and surfactant molecules. Ammonia and nitrate ions were checked by means of concentrated hydrochloric acid and by using the brown ring test, respectively. Finally the precipitate was washed with ethanol to facilitate the drying process. The obtained white precipitate was oven dried at 50 °C and calcined at 900 °C for 1 h. Methods. SANS measurements were performed at the Rutherford Appleton Laboratory, on the ISIS pulsed neutron source, Didcot (Oxford), U.K., by using the time-of-flight SANS instrument LOQ.42 (41) Giannakas, A. E.; Vaimakis, T. C.; Ladavos, A. K.; Trikalitis, P. N.; Pomonis, P. J. J. Colloid Interface Sci. 2003, 259, 244.
Langmuir, Vol. 23, No. 7, 2007 3949 Samples were prepared by using D2O instead of H2O to enhance the contrast (i.e., difference in the scattering length densities among various regions in the system). They were contained in fused silica cells of 2 mm path length and placed 4.3 m from the main twodimensional detector. The temperature was kept at 25.0 ( 0.2 °C by an external circulating oil bath. Data from the main twodimensional detector43 were averaged around anular rings before correction for transmission, detector efficiency, and monitor response. Sample transmission was corrected for the wavelength dependence. Simultaneous neutron diffraction data were combined in the wavelength range 2-10 Å to give net intensities in the Q range between 0.01 and 0.23 Å-1 (the scattering vector, Q, is 4π(sin θ)/λ, where 2θ is the scattering angle). The net intensities were converted to absolute differential scattering cross sections per unit sample volume (dΣ(Q)/dΩ, cm-1) by comparison with precalibrated secondary standards.44,45 Scattering from the cell blank was subtracted from that of each sample. The sample incoherent scattering and the instrumental background were evaluated from the slope of the linear regression of Q4I(Q) vs Q4, in the Porod region of the scattering curve, and subtracted from each data point.46 Details on data reduction are reported elsewhere.43 WAXS measurements were performed by a Philips PW 1050/39 diffractometer in the Bragg-Brentano geometry using Ni-filtered Cu KR radiation (λ ) 1.54178 Å). The X-ray generator worked at a power of 40 kV and 30 mA, and the resolution of the instrument (divergent and antiscatter slits of 0.5°) was determined using R-SiO2 and R-Al2O3 standards free from the effect of reduced crystallite size and lattice defects. TEM analysis has been performed by using a JEOL 2010 operating at an accelerating voltage of 200 kV. The Nd:YAG powder was dispersed in ethanol, and to ensure a more homogeneous dispersion, the suspension was sonicated. A small drop was deposited on a copper grid of 300 mesh, and after complete solvent evaporation the copper grid was introduced into the TEM chamber analysis. Specific surface area measurements were carried out using physical adsorption of nitrogen by means of a Micrometrics 2360 surface area analyzer. Density. The powder density was determined by performing accurate measurements of weight and volume by means of an AccuPyc 1330 helium pycnometer. The sample was weighted on an analytical balance, and its volume was obtained from the difference of the volumes of the test chamber filled with helium and helium plus the sample. Photoluminescence. Spectra in the near-infrared range were obtained at room temperature, excitation being with the 488.0 nm radiation of a Spectra Physics argon laser, model 2017. The scattered signal was analyzed by a 1/2 m monochromator equipped with a 150 lines/mm grating and a CCD detector (Spectrum One, Jobin-Yvon). A fiber optic probe was employed. The resolution of the emission spectra was about 0.5 nm. The Nd3+ ion luminescence decay curve was measured under excitation with the third harmonic radiation of a Nd:YAG pulsed laser (355 nm). The signal was detected using the above-mentioned monochromator and a cooled GaAs photomultiplier (Hamamatsu).
Results and Discussion Microemulsion Characterization. The scattering intensity vs Q obtained for the D2O/CTAB/1-butanol/n-heptane system, at the composition reported above, is shown in Figure 1A. The curve shape and the peak position indicate the occurrence in the system of regions spatially organized. The scattering cross sections for the microemulsion containing neodimium-yttrium-alumi(42) Heenann, R. K.; Penfold, J.; King, S. M. J. Appl. Crystallogr. 1997, 30, 1140. (43) King, S. M. In Modern Techniques for Polymer Characterisation; Pethrick, R. A., Dawkins, J. V., Eds.; J. Wiley & Sons: New York, 1999. (44) Wignall, G. D.; Bates, F. S. J. Appl. Crystallogr. 1987, 20, 28. (45) Russel, T. P.; Lin, J. S.; Spooner, S.; Wignall, G. D. J. Appl. Crystallogr. 1988, 21, 629. (46) Porod, G. Kolloid-Z. 1951, 124, 83.
3950 Langmuir, Vol. 23, No. 7, 2007
Caponetti et al.
Figure 3. TEM micrographs of Nd:YAG nanopowder calcined at 900 °C for 1 + 1 h. Table 1. Values of Floating and Derived Parameters in the Fitting Procedure of SANS Dataa aqueous phase no salt nitrates a
103a1
c1
c2
ξ (Å)
d (Å)
ξ/d
fa
χ
0.30(5) 0.26(7)
-0.229(7) -0.234(5)
72(2) 91(1)
66.4 70.5
148 163
0.45 0.43
-0.77 -0.76
1.4 1.5
The error on the last digit is given in parentheses.
num nitrates and ammonium hydroxide are reported in parts B and C, respectively, of Figure 1. The presence of nitrates causes a strong intensity increase and a small peak position displacement toward lower Q values, while the presence of ammonium hydroxide causes a marked variation of the curve shape, clearly indicating a change in the system structure. SANS data, related to the system without reactants, were modeled assuming a distribution of water droplets surrounded by a mixed surfactant/alcohol layer and dispersed in heptane, i.e., the typical model used for the L2 phase of a w/o microemulsion.47 Notwithstanding the variety of models tested, i.e., monodisperse and polydisperse spheres and ellipsoidal droplets interacting via hard48 or soft49 sphere potential, the experimental data were not reproduced. As an example, the curve obtained by using the model of ellipsoidal particles, interacting via hard-sphere potential, is reported as a dashed line in Figure 1A. Hence, the functional form of the scattering formula developed by Teubner and Strey51 on the basis of a phenomenological model was checked. The scattering intensity has been calculated by applying the following equation:
8πc2〈η2〉/ξ dΣ(Q) ) dΩ a2 + c1Q2 + c2Q4
(1)
where 〈η2〉 ) φwφo〈∆F2〉, φw being the volume fraction of the aqueous medium, φo the volume fraction of the oil medium, and ∆F the difference in scattering length densities between the aqueous and oil media. Q is the scattering vector previously defined, and ξ is the correlation length. a2, c1, and c2 are the coefficients appearing in the phenomenological Landau free energy expression obtained from an order parameter expansion of the free energy density and are related to the stability of the system. They are sufficient to describe a variety of structures, (47) Compere, A. L.; Griffith, W. L.; Johnson, J. S., Jr.; Caponetti, E.; Chillura Martino, D.; Triolo, R. J. Phys. Chem. 1997, 101, 7139 and references therein. (48) Ashcroft, N. W.; Lekner, J. Phys. ReV. 1966, 145, 83. (49) Blum, L.; Stell, G. J. Chem. Phys. 1979, 71, 42 and 1980, 72, 2212. (50) Caponetti, E.; Chillura Martino, D.; Saladino, M. L. Structure of the CTAB/buthanol/water/n-heptane quaternary microemulsion. A SANS Study. To be submitted for publication. (51) Teubner, M.; Strey, R. J. Chem. Phys. 1987, 87, 3195.
going from binary systems of two partially miscible liquids, to spinodal decomposition as well as isotropic multicomponent liquids.46 ξ and the second length scale characteristic of a microemulsion, d, i.e., the quasi-periodic repeat distance between polar and nonpolar regions, are both related to the coefficients a2, c1, and c2 by means of the following expressions:
ξ)
[( ) ] [( ) ] 1 a2 2 c2
d ) 2π
1/2
1 a2 2 c2
+
1/2
c1 4c2
-
-1/2
c1 4c2
(2)
-1/2
(3)
The scattering intensity computed by means of eq 1, by allowing a1, c1, and c2 to vary, is reported as a continuous line in Figure 1A. It results in good agreement with the experimental data (points). Values of floating parameters and of those derived by means of eqs 2 and 3 are reported in Table 1. The positive sign of a2 and c2 and the negative sign of c1 are as expected.31 To get information about the structure of the system, the ratio ξ/d and the amphiphilicity factor, fa, have been evaluated. The ratio ξ/d was introduced by Teubner and Strey51 as a guessing parameter to explain and foresee the appearance of a lamellar phase on increasing the surfactant concentration. The ξ/d value ranges from 0.35 up to 0.76 for a variety of systems reported.51 fa is related to the stability of the system, in terms of the coefficients a2, c1, and c2:52-53
fa ) c1/(4c2a2)1/2
(4)
Its value ranges from +1 for a completely disordered phase to -1 for a lamellar phase. The obtained values of both parameters are reported in Table 1. The ξ/d value, 0.45, is in good agreement with the value reported for a microemulsion whose bicontinuous structure is supported by diffusion NMR data.51 fa is -0.77, confirming that the system is a bicontinuous one, but it is not far from the boundary line of the lamellar phase. (52) Schubert, K. V.; Strey, R. J. Chem. Phys. 1991, 95, 8532. (53) Schubert, K. V.; Strey, R.; Kline, S. R.; Kaler, E. W. J. Chem. Phys. 1994, 101, 5343.
Preparation of Nd:YAG Nanopowder
Langmuir, Vol. 23, No. 7, 2007 3951
Figure 4. Photoluminescence spectrum of Nd:YAG nanopowder calcined at 900 °C for 1 + 1 h.
The bicontinuous model has also been used to analyze SANS data related to the system containing neodymium, yttrium, and aluminum nitrates. The results are reported in Figure 1B. As can be seen the agreement between the computed intensity (line) and experimental data (points) reported is fairly good. Values of floating parameters and of those derived are reported in Table 1. The presence of nitrates induces an increase in both the length scale parameters (ξ, d). The bigger increase observed for d suggests that nitrates are more efficient to move away layers of surfactant than to induce a longer length scale correlation. The ξ/d value, 0.43, again, falls within the range of values characteristic of bicontinuous microemulsions. The surfactant concentration being constant in the present case, the slight decrease of its value could be interpreted by assuming that nitrates stabilize the bicontinuous structure. This is confirmed by the value of fa, -0.76, very close to the previous one, indicating that the salt addition slightly affects the degree of order of the system and, furthermore, is not sufficient to promote the phase evolution toward the lamellar phase. The attempt to reproduce data related to the microemulsion containing ammonium hydroxide via the bicontinuous model is useless as can be seen by the comparison among experimental data (points) and the computed intensity, reported as a dotted line in Figure 1C. Considering that the neat system is constituted by a bicontinuous structure lying near the borderline of the lamellar phase region, our guess was that the presence of ammonia could induce the phase transition. On this basis, the model proposed by Nallet et al.54 for a lamellar phase was used to fit the experimental data. In such a model, the following equation applies for the scattered intensity:
dΣ(Q) B A + ) 2 2 dΩ Q ξp + 1 (Q - Qmax)2ξl2 + 1
(5)
It consists of two parts: the former represents the small-angle diffuse scattering arising from the amphiphile concentration fluctuation, and the latter represents the quasi-Bragg peak due to the stacking of the lamellar membranes. A and B correspond to the scattering intensities from the diffuse scattering and the quasi-Bragg scattering, ξp is the correlation length of the concentration fluctuation of the amphiphile, ξl is the spatial correlation length, and Qmax is the Q value corresponding to the peak intensity. The computed intensity by means of eq 5 by allowing A, B, ξp, ξl, and Qmax to vary is reported as a continuous line in Figure (54) Nallet, F.; Roux, D.; Milner, S. T. J. Phys. (Paris) 1990, 51, 2333. (55) Inorganic Crystal Structure Database. http://icsdweb.FIZ-Karlsruhe.de.
1C. The fair agreement between the computed intensity and experimental data indicates that the lamellar phase model is able to describe the structure of the microemulsion in the presence of ammonium hydroxide. The A and B values were 82 ( 2 and 131 ( 2, respectively. The two correlation lengths ξp and ξl were 42 ( 2 and 188 ( 5 Å, respectively. Qmax was 0.030 ( 0.001, from which d ) 158 ( 5 Å (d ) 2π/Qmax) was evaluated. The above results clearly indicate that, even if the size domains were almost unchanged, and a slight effect on the d values with composition was observed, the microemulsion evolves from a bicontinuous structure toward a lamellar one, as a consequence of the variation in the chemical composition of the aqueous phase. Probably, the presence of ammonium hydroxide is more effective in decreasing the amphiphilicity factor than nitrate salts, thus stabilizing the lamellar structure. Nd:YAG Nanopowder Characterization. WAXS. From a close inspection of the spectrum of calcined Nd:YAG powder, reported in Figure 2 (lower spectrum), it is clearly seen that the YAG phase is accompanied by the hexagonal form of YAlO3 (YAH phase), whose peaks are highlighted by open circles. The YAH phase is one of the intermediate phases appearing during the calcination process; other possible intermediate phases are the orthorhombic form of Y4Al2O9 (YAM phase), Y2O3, and Al2O3.16,56-58 The different phases were identified by using the JCPDS powder diffraction files.55 Even if the YAG phase is the predominant one, the YAH phase is not negligible; for this reason as suggested by other studies, a 1 h additional treatment at 900 °C has been performed on the same sample. The relative spectrum is reported in Figure 2 (upper spectrum). The amount of the YAH phase is appreciably reduced, indicating that the second treatment results in an almost complete conversion of YAH into the YAG phase. The exact hkl peaks positions were determinated by fitting each peak by means of the Pearson VII function. The lattice parameter (a) was, then, computed by the Celsize program59 using the formula valid for a cubic lattice:
a ) dhkl(h2 + k2 + l2)1/2
(6)
The refined value of the lattice parameter turned out to be 12.057 ( 0.001 Å. The comparison with the lattice parameter of the pure YAG structure55 (a ) 12.053 ( 0.001 Å) indicates a small increase due to the partial substitution of Y3+ ion sites with Nd3+. TEM. Three representative TEM micrographs of Nd:YAG nanopowder calcined at 900 °C for 1 + 1 h are reported in Figure 3. The powder is constituted by very fine spherical nanoparticles, whose size is about 20 nm (Figure 3B), showing a certain tendency to agglomerate (Figure 3A). Figure 3C shows the contact region of two particles. The nanoparticle size is smaller than that of materials obtained by other synthesis methods.15-18,36 It is noteworthy that single particles were not detected in the sample prepared by the coprecipitation method,36 where the nanopowder was mainly composed of aggregates of primary particles. Since the aggregation degree has a dramatic effect on the sintering process60 and on the possibility of their homogeneous dispersion in an (56) Hess, N. J.; Maupin, G. D.; Chick, L. A.; Sunberg, D. S.; McCreedy, D. E.; Armstrong, T. R. J. Mater. Sci. 1994, 29, 1873. (57) Hay, R. S. J. Mater. Res. 1993, 8, 578. (58) Veitch, C. D. J. Mater. Sci. 1991, 26, 6527. (59) Hay, D. Celsize Software Program; CSIRO Division of Material Science and Technology: Clayton, Australia. (60) Li, J.-G.; Ikegami, T.; Mori, T. Acta Mater. 2004, 52, 2221.
3952 Langmuir, Vol. 23, No. 7, 2007
Caponetti et al.
appropriate medium, it follows that the powder obtained in the microheterogeneous system should be suitable material for producing ceramics and polymer-nanoparticle composites, starting materials for the production of lasers and LEDs, respectively. Specific Surface Area. The specific surface area (SBET) of the powder calcined at 900 °C for 1 + 1 h has been calculated from nitrogen adsorption data using the BET equation.61SBET was 44.6 ( 0.1 m2 g-1, 1 order of magnitude greater than that of the same material obtained by the coprecipitation method (3.15 ( 0.06 m2 g-1), confirming that the particles are smaller and/or their degree of aggregation is lower. Once known by TEM that the nanoparticles are roughly spherical, the apparent equivalent diameter (dBET) can be estimated from the specific surface area and the density by using the expression
dBET ) 6/(FSBET)
(7)
F, determined using accurate measurements of volume and weight, was 2.69 g cm-3. The dBET value was 50 nm, higher than that evaluated by TEM, indicating that some regions of the sample are not accessible to the gas as a consequence of a close contact between particles. Photoluminescence. The emission spectrum of Nd:YAG nanopowder is reported in Figure 4. The emission bands were assigned to transitions from the thermalized Stark levels of the 4F3/2 excited state to those of the 4I9/2 ground state. The energies of the emission transitions correspond to those reported for a Nd3+-doped YAG crystal,62 confirming that the dopant Nd3+ ions have entered the YAG crystalline lattice. The lifetime value obtained from the luminescence decay curves was 286 ( 10 µs, higher than that of the single crystal (255 µs) and that of the nanopowders obtained by the coprecipitation method (75 µs).36 (61) Brunauer, S.; Emmett, P. H.; Teller, E. J. Am. Chem. Soc. 1938, 60, 309. (62) Kaminskii, A. A. Laser Crystal, 2nd ed.; Springer-Verlag: Berlin, 1990.
Conclusion The synthesis of yttrium aluminum garnet doped with neodimium ions has been performed by coprecipitating the precursor hydroxides in the quaternary system constituted by water, cetyltrimethylammonium bromide, 1-butanol, and nheptane. At the examined compositions, the structure of the microemulsion, as emerged from SANS data analysis, is a quite ordered bicontinuous structure. It does not change when nitrates are added, but it transforms into a lamellar structure when ammonium hydroxide is added. Since the reaction takes place when the microemulsion containing ammonium hydroxide is dropwise added to the microemulsion containing nitrates, we think that the hydroxide coprecipitation takes place in a system that continuously evolves from a bicontinuous to a lamellar phase. After calcinations at 900 °C for 1 + 1 h, the obtained Nd:YAG nanopowders have a bigger surface area and enhanced optical properties compared to the same material produced by the traditional coprecipitation method. It is constituted by aggregates of nanoparticles and single nanoparticles having a smaller size. It can be concluded that the bicontinuous/lamellar phase inhibits the indefinite growth of clusters, demonstrating that isolated aqueous pools (L2 phase) are not strictly required to limit the nanoparticle growth. The presence of surfactant and the “confined environments” in which reactants come in contact seems to be determining. The synthetic protocol that has been followed is very simple. It allows high yield and can be scaled up without particular problems. It seems to be very promising to obtain starting materials for the production of ceramics and for the realization of polymernanoparticle composites, products used for laser and LED construction, respectively. Acknowledgment. Thanks are due to the Ministero dell’Universita` e della Ricerca and to Universita` di Palermo for financial funding. We acknowledge Prof. M. Bettinelli and Dr. A. Speghini (Universita` di Verona, Italy) for photolumiscence measurements and Prof. S. Enzo (Universita` di Sassari, Italy) for useful discussion on WAXS spectrum interpretation. LA0625906
