- Email: [email protected]
Flame aerosol synthesis of phase-pure monoclinic Y2O3 particles via particle size control
Powder Technology 191 (2009) 231–234
Contents lists available at ScienceDirect
Powder Technology j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / p o w t e c
Short communication
Flame aerosol synthesis of phase-pure monoclinic Y2O3 particles via particle size control Bing Guo ⁎, Mallika Mukundan, Hoon Yim Department of Mechanical Engineering, Texas A&M University, College Station, Texas, USA
a r t i c l e
i n f o
Article history: Received 23 August 2008 Received in revised form 4 October 2008 Accepted 3 November 2008 Available online 18 November 2008 Keywords: Flame aerosol synthesis Impactor Size effect Crystal structure Polymorphism
a b s t r a c t In this study, for the first time, a particle size effect on crystal structure of Y2O3 particles was exploited to synthesize phase-pure monoclinic Y2O3 particles. In the synthesis process, a precursor aerosol consisting of H2 fuel gas and precursor droplets passed through an impactor before it entered a flame to form yttria particles. A round-jet impactor was used to remove the large precursor droplets, so that the product Y2O3 particles were all smaller than a critical size of approximately 1.5 µm. Due to the particle size effect on crystal structure, the Y2O3 particles thus obtained were essentially phase-pure with the monoclinic structure. The result shows that, by using an impactor to alter the particle size distribution, it is possible to control the crystal structure of Y2O3 particles while maintaining relatively high synthesis yield. © 2008 Elsevier B.V. All rights reserved.
1. Introduction Y2O3 is a material that finds applications in phosphors, catalysts and optical window materials. Y2O3 has multiple crystal structures with differing properties. Mainly two phases of Y2O3 are possible from regular synthesis processes, namely the C-type cubic phase and the B-type monoclinic phase [1]. The two crystal structures have significantly different thermophysical and optical properties. For example, the density of monoclinic Y2O3 is significantly higher than that of the cubic phase [2]; the fluorescence properties of Eu-doped Y2O3, an important phosphor material, are strongly dependent upon the Y2O3 crystal structure [3]. So far the application for monoclinic Y2O3 particles has not been extensively explored. One apparent reason is simply that, until now there has not been a feasible synthesis method to produce this material in large quantities. Nevertheless, the monoclinic structure is an important phase for Y2O3 and other rare earth sesquioxides [6]. A number of researchers have studied the synthesis of monoclinic Y2O3 using various methods. A summary of the reported methods for synthesizing monoclinic Y2O3 particles is given in Table 1. High-pressure processes require special equipment and are not capable of producing nanoparticles or microparticles [1]. Furnace-based heating–condensation methods require two
⁎ Corresponding author. Department of Mechanical Engineering, Texas A&M University, 3123 TAMU, College Station, TX 77843, USA. Tel.: +1 979 845 8450; fax: +1 979 845 3081. E-mail address: [email protected] (B. Guo). 0032-5910/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.powtec.2008.11.003
steps, can only produce monoclinic Y2O3 particles b10 nm, and are not suited for continuous synthesis of significant quantities of material [2]. In a previous study, a flame process with a gas-phase precursor was used to synthesize monoclinic Y2O3 particles, but the particle size was limited to below 90 nm due to the gas-phase precursor. The batch mode precursor loading also makes it difficult to achieve continuous synthesis with that method. In addition, the precursor used in that method is costly [4]. More recently, a flame spray pyrolysis (FSP) method was used to synthesize monoclinic Y2O3 particles with diameters up into the micrometer range. However, the polydisperse Y2O3 particles had mixed cubic and monoclinic phases. A critical particle diameter of approximately 1.5 µm was found. At the critical diameter, the probability was 50% for a particle to be either cubic or monoclinic. Particles significantly smaller than the critical diameter were all monoclinic, while those significantly larger were all cubic. [5]. To explore the potential applications for monoclinic Y2O3, one must first be able to synthesize phase-pure Y2O3 in sufficient quantities. This synthesis capability is also key to studying the interplay between surface energy and polymorphism [7]. The relationship between surface energy and polymorphism is a topic of profound importance in materials formation, especially on the nanometer scale [8]. This work was motivated by the above-mentioned reasons. The basic operating principle used in this study was that, if all the Y2O3 particles produced from the flame synthesis process were smaller than the critical size, then they would all have the monoclinic structure. Therefore phase-pure monoclinic Y2O3 particles may be generated via controlling the particle size. Herein we report the experimental methods for particle size control and the respective results. In particular, we describe the design and the successful use of a real impactor for achieving particle size control. To the
232
B. Guo et al. / Powder Technology 191 (2009) 231–234
Table 1 Summary of methods for synthesizing monoclinic Y2O3 particles Author
Method
Maximum process temperature (˚C)
Critical particle sizea
Yield
Hoekstra [1] Krauss et al. [2] Guo et al. [4]
High pressure diamond anvil Evaporation–condensation Flame aerosol process with gas-phase precursor Flame aerosol process with droplet precursor (flame spray pyrolysis)
~ 1000 ˚C 350 ˚C ~ 2700 ˚C
∞ (Bulk sample) ~ 10 nm Not observed (All particles monoclinic largest ~ 90 nm) ~ 1.5 µm (Mixed with larger Y2O3 particles that were cubic)
Unspecified ~ 150 mg/batch (inferred from sample description) 40–120 mg/h (batch mode; each run approximately 15 min) 200 mg /h (0.65 M precursor solution) and 10 mg/h (0.026M precursor solution)
Guo and Luo [5]
a
~ 2700 ˚C
Critical size is the size below which the Y2O3 particles had the monoclinic structure, and those larger had the cubic structure.
best of our knowledge, this is the first report of a real impactor being used in flame spray pyrolysis to obtain phase-pure product particles. 2. Experimental methods 2.1. Flame spray pyrolysis apparatus The flame spray pyrolysis apparatus is schematically shown in Fig. 1. It is similar to the apparatus used in a previous work [5], except that an optional impactor was incorporated in the apparatus in this study. The inner/outer diameters of the burner nozzle are 1.6 mm and 9.5 mm, respectively. The apparatus consisted of a 1.7 MHz atomizer, an atomization vessel, a furnace and a burner. The furnace and the burner have been described in detail elsewhere [4]. When the optional impactor was used, the precursor aerosol had to flow through the impactor first before entering the furnace and burner. The impactor removed the large droplets from the precursor aerosol. The precursor aerosol formed a self-sustained steady-state aerosol flame at the top of the burner. A co-flowing oxidant stream supported the flame. The furnace heated precursor aerosol to maintain a
sufficiently high flame temperature. The precursor aerosol underwent chemical reactions in the flame and became a Y2O3 aerosol. The postflame aerosol containing Y2O3 particles was drawn into a sampling tube by vacuum and the particles were collected on an alumina membrane filter (Whatman Inc., NJ). The synthesis apparatus was operated at atmospheric pressure. H2 was used as the fuel gas at a flow rate of 1 SLM (standard liter per minute). Pure O2 at 6 SLM was used as the oxidant stream to support the flame. The flame length was approximately 5 cm. The precursor solution was prepared by dissolving yttrium nitrate hexahydrate (chemical formula Y (NO3)3·6H2O, 99.9%, Alfa Aesar, Ward Hill, MA) in Nanopure® water (Barnstead, Dubuque, IO). The concentrations of the precursor solution used in this study were from 0.026 to 0.65 M. 2.2. Impactor design The objective of the impactor design was to ensure removal efficiency greater than 80% for 6-µm droplets. With a 0.65-M precursor solution, a 6-µm droplet would produce a 1.5-µm Y2O3 particle, assuming that one droplet becomes one final Y2O3 particle. Based on mass conservation one can readily infer the relation between a precursor droplet diameter and the resultant Y2O3 particle diameter, knowing the precursor concentration. This relation neglects the evaporation of Y2O3 particles and coagulation between Y2O3 particles. These assumptions will be discussed later. The primary dimensions of the impactor were determined using the relations given by Marple and Willeke [9]. The design gas was H2 at 1 SLM. The droplet density was 1120 kg/m3 (measured density for the 0.65 M precursor solution). The impactor had a round nozzle with an inner diameter of 2.87 mm. The nozzle inner diameter was selected by trial and error, so that the precursor droplets had the proper Stokes number, and hence desired removal efficiency in the impactor. The relation between removal efficiency and Stokes number (and the Reynolds number to a lesser degree) was found in the paper by Marple and Willeke [9]. The nozzleimpaction plate distance was 3.30 mm, selected based on the empirical relation for circular jet impactors [10]. Estimated removal (collection) efficiency and the corresponding Y2O3 particle diameter for several droplet sizes are given in Table 2. A schematic drawing of the impactor is shown in Fig. 2. 2.3. Computational fluid dynamics simulation for the impactor Computational fluid dynamic (CFD) simulation was used to verify the droplet removal efficiency for the impactor. The flow field through the impactor was simulated using FLUENT 6.2 (Fluent Inc., Lebanon, Table 2 Estimated removal efficiency and the corresponding Y2O3 particle diameter for several droplet sizes
Fig. 1. Schematic of flame spray pyrolysis apparatus for Y2O3 synthesis.
Droplet diameter, 0.65-M solution (µm) Stokes number Estimated efficiency of removal by impactor [9] Resultant Y2O3 particle diameter (µm)
4 0.20 b 20% 1.0
6 0.45 N 80% 1.5
6.8 0.58 N 90% 1.7
B. Guo et al. / Powder Technology 191 (2009) 231–234
233
This agrees with the expected performance of the impactor given in the Experimental Methods section. 3.2. Simulated particle removal efficiency Fig. 5 shows the particle removal efficiency calculated based on CFD simulation. It agrees well with the estimated efficiency shown in Table 2. The simulation shows impactor removal efficiency to be 100% for droplets larger than 7 µm. This is in agreement with the experimentally measured Y2O3 particle size distribution with impactor, in which particles larger than 1.5 µm are absent. 3.3. Crystal structure of Y2O3 particles
Fig. 2. A schematic drawing of the round jet impactor for removing large precursor droplets.
NH) with the appropriate models. The bulk fluid (continuous phase) flow field was solved using the Eulerian approach. After solving the flow field, the Discrete Phase Formulation was used to simulate particles trajectories based on the Lagrangian approach. The particle trajectories were calculated with the assumption that the particles neither influence the flow field nor each other. Trajectories for particle sizes in the range of 1–10 µm were generated. For each size, 500 trajectories were simulated with initial particle locations randomly placed at the inlet of the flow field. Particle removal efficiency for each size was then calculated based on the simulated particle trajectories. 3. Results
The XRD results in Fig. 6 show that with the impactor, the Y2O3 particles were phase-pure monoclinic. Without the impactor, the Y2O3 contained a significant amount of the cubic phase. This agrees with the fact that without the impactor a significant number of the Y2O3 particles were larger than the critical size, while the particles with the impactor were all smaller than the critical size. 4. Discussion 4.1. Size control via precursor concentration control As found in the previous study, when the precursor solution concentration was lowered to 0.026 M, the product Y2O3 particles were phase-pure monoclinic. Thus one could generate phase-pure monoclinic Y2O3 particles by lowering the precursor concentration [5]. However, that required the concentration to decrease by more than a factor of twenty. Because the particle yield is proportional to the precursor concentration, the low precursor concentration leads to
3.1. Effectiveness of impactor Typical TEM images of particles synthesized without the impactor and with the impactor are shown in Fig. 3. As can be seen in the TEM image, without the impactor, Y2O3 particles as large as 2 µm are present; while with the impactor, such large particles are absent. The effect of the impactor became evident when the particle size distribution was obtained by surveying at least 300 particles in the TEM images. The particle size distributions are shown in Fig. 4 for representative samples with and without the impactor. Without the impactor, a significant fraction of the particles are larger than 1.5 µm, the critical particle diameter. With the impactor, the largest particles are approximately 1.5 µm in diameter. Particles larger than 1.5 µm were essentially absent from the samples generated with the impactor. This suggests that the impactor had indeed removed precursor droplets larger than 6 µm (see Table 2 for size correlation).
Fig. 3. TEM images of Y2O3 particles synthesized (A) without and (B) with the impactor to remove large precursor droplets.
Fig. 4. Size distribution of Y2O3 synthesized without and with the impactor.
234
B. Guo et al. / Powder Technology 191 (2009) 231–234 Table 3 Effect of precursor concentration and use of impactor on synthesis yield
Fig. 5. Particle removal efficiency based on CFD simulation.
very low mass yield, as shown in Table 1. On the other hand, using the impactor, only the largest particles are stripped off the particle size distribution, hence a relatively high yield can be maintained while ensuring the particles to be phase-pure monoclinic. A comparison of the Y2O3 particle yield is given in Table 3. The flame used in this study was a laminar, quiet and steady flame. If particle disintegration and coagulation are negligible in the flame, then one precursor droplet becomes one product particle. In other flame spray pyrolysis processes, an air-blast type atomizer is integrated with the burner, and the precursor solvent is also a fuel [11]. The flame in those processes is a turbulent reacting jet, in which strong droplet/particle disintegration exists. One can find out that in such a process the onedroplet-one-particle relation does not apply [12].
Y(NO3)3·6H2O concentration (M)
0.65
Yield of Y2O3 particles (mg/h) Crystal structure
200 Mixed
0.65 (with impactor) 140 Phase-pure monoclinic
0.026 [5] 10 Phase-pure monoclinic
number flow rate may be calculated. Then the particle concentration in the flame was estimated to be 2.5 × 1012 1/m3, assuming it was the same as the precursor droplet number concentration in the fuel gas. Assuming the gas was air at 2700 K, for 800-nm particles, the estimated coagulation rate was about 1.5 × 1010 1/m3 s [10]. The particles’ residence time in the flame was in the order of 1 ms [4]. Within this time, coagulation between particles was negligible. In other words, vast majority of the Y2O3 particles would not experience significant coagulation due to the relatively low particle concentration and the short residence time in the flame. 5. Summary and conclusion By incorporating an impactor in a flame spray pyrolysis apparatus, the product Y2O3 particles were kept below 1.5 µm in diameter. The particles were all smaller than the critical diameter that determines the crystal structure of Y2O3 in this flame synthesis process. As a result, the Y2O3 particles were phase pure with the monoclinic crystal structure. The result from this study shows that, in spray pyrolysis where a particle size effect on crystal structure exists, one can use a simple impactor to obtain phasepure particles of one crystal structure. Acknowledgement
4.2. Negligible Y2O3 particle coagulation In this study it was assumed that each precursor droplet became one Y2O3 particle. Then we assumed that by removing those large droplets that would make Y2O3 particles larger than the 1.5 µm, we could prevent the formation of cubic Y2O3 particles. However, in an aerosol smaller particles coagulate to form larger particles. Thus particles larger than 1.5 µm may be formed from smaller particles. Therefore, the coagulation rate was estimated for the Y2O3 particles in the flame. Based on the precursor solution atomization rate (5 mL/ min) and an estimated average droplet diameter of 4 µm, the particle
Fig. 6. XRD of Y2O3 synthesized without and with the impactor; filled diamonds mark major distinctive peaks of cubic Y2O3.
Financial support for this work was provided by Texas Engineering Experiment Station and Texas A&M University. The Microscopy and Imaging Center at Texas A&M University made available the electron microscope used in this study.
References [1] H.R. Hoekstra, Phase relationships in the rare earth sesquioxides at high pressure, Inorganic Chemistry 5 (5) (May 1966) 754–757. [2] W. Krauss, R. Birringer, Metastable phases synthesized by inert-gas-condensation, Nanostructured Materials 9 (1–8) (1997) 109–112. [3] X. Qin, Y.G. Ju, S. Bernhard, N. Yao, Flame synthesis of Y2O3: Eu nanophosphors using ethanol as precursor solvents, Journal of Materials Research 20 (11) (Nov 2005) 2960–2968. [4] B. Guo, A. Harvey, S.H. Risbud, I.M. Kennedy, The formation of cubic and monoclinic Y2O3 nanoparticles in a gas-phase flame process, Philosophical Magazine Letters 86 (7) (Jul 2006) 457–467. [5] B. Guo, Z.P. Luo, Particle size effect on the crystal structure of Y2O3 particles formed in a flame aerosol process, Journal of the American Ceramic Society 91 (5) (2008) 1653–1658. [6] B. Guo, A.S. Harvey, J. Neil, I.M. Kennedy, A. Navrotsky, S.H. Risbud, Atmospheric pressure synthesis of heavy rare earth sesquioxides nanoparticles of the uncommon monoclinic phase, Journal of the American Ceramic Society 90 (11) (2007) 3683–3686. [7] P. Zhang, A. Navrotsky, B. Guo, I. Kennedy, A.N. Clark, C. Lesher, et al., Energetics of cubic and monoclinic yttrium oxide polymorphs: phase transitions, surface enthalpies, and stability at the nanoscale, Journal of Physical Chemistry C 112 (2008) 932–938. [8] A. Navrotsky, Energetics of nanoparticle oxides: interplay between surface energy and polymorphism, Geochemical Transactions 4 (Nov 18 2003) 34–37. [9] V.A. Marple, K. Willeke, Impactor design, Atmospheric Environment 10 (10) (1976) 891–896. [10] W.C. Hinds, Aerosol Technology: Properties, Behavior, and Measurement of Airborne Particles, 2nd edWiley, New York, 1999. [11] L. Madler, H.K. Kammler, R. Mueller, S.E. Pratsinis, Controlled synthesis of nanostructured particles by flame spray pyrolysis, J Aerosol Sci 33 (2) (Feb 2002) 369–389. [12] D. Dosev, B. Guo, I.M. Kennedy, Photoluminescence of Eu3+: Y2O3 as an indication of crystal structure and particle size in nanoparticles synthesized by flame spray pyrolysis, J Aerosol Sci 37 (3) (Mar 2006) 402–412.
Contents lists available at ScienceDirect
Powder Technology j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / p o w t e c
Short communication
Flame aerosol synthesis of phase-pure monoclinic Y2O3 particles via particle size control Bing Guo ⁎, Mallika Mukundan, Hoon Yim Department of Mechanical Engineering, Texas A&M University, College Station, Texas, USA
a r t i c l e
i n f o
Article history: Received 23 August 2008 Received in revised form 4 October 2008 Accepted 3 November 2008 Available online 18 November 2008 Keywords: Flame aerosol synthesis Impactor Size effect Crystal structure Polymorphism
a b s t r a c t In this study, for the first time, a particle size effect on crystal structure of Y2O3 particles was exploited to synthesize phase-pure monoclinic Y2O3 particles. In the synthesis process, a precursor aerosol consisting of H2 fuel gas and precursor droplets passed through an impactor before it entered a flame to form yttria particles. A round-jet impactor was used to remove the large precursor droplets, so that the product Y2O3 particles were all smaller than a critical size of approximately 1.5 µm. Due to the particle size effect on crystal structure, the Y2O3 particles thus obtained were essentially phase-pure with the monoclinic structure. The result shows that, by using an impactor to alter the particle size distribution, it is possible to control the crystal structure of Y2O3 particles while maintaining relatively high synthesis yield. © 2008 Elsevier B.V. All rights reserved.
1. Introduction Y2O3 is a material that finds applications in phosphors, catalysts and optical window materials. Y2O3 has multiple crystal structures with differing properties. Mainly two phases of Y2O3 are possible from regular synthesis processes, namely the C-type cubic phase and the B-type monoclinic phase [1]. The two crystal structures have significantly different thermophysical and optical properties. For example, the density of monoclinic Y2O3 is significantly higher than that of the cubic phase [2]; the fluorescence properties of Eu-doped Y2O3, an important phosphor material, are strongly dependent upon the Y2O3 crystal structure [3]. So far the application for monoclinic Y2O3 particles has not been extensively explored. One apparent reason is simply that, until now there has not been a feasible synthesis method to produce this material in large quantities. Nevertheless, the monoclinic structure is an important phase for Y2O3 and other rare earth sesquioxides [6]. A number of researchers have studied the synthesis of monoclinic Y2O3 using various methods. A summary of the reported methods for synthesizing monoclinic Y2O3 particles is given in Table 1. High-pressure processes require special equipment and are not capable of producing nanoparticles or microparticles [1]. Furnace-based heating–condensation methods require two
⁎ Corresponding author. Department of Mechanical Engineering, Texas A&M University, 3123 TAMU, College Station, TX 77843, USA. Tel.: +1 979 845 8450; fax: +1 979 845 3081. E-mail address: [email protected] (B. Guo). 0032-5910/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.powtec.2008.11.003
steps, can only produce monoclinic Y2O3 particles b10 nm, and are not suited for continuous synthesis of significant quantities of material [2]. In a previous study, a flame process with a gas-phase precursor was used to synthesize monoclinic Y2O3 particles, but the particle size was limited to below 90 nm due to the gas-phase precursor. The batch mode precursor loading also makes it difficult to achieve continuous synthesis with that method. In addition, the precursor used in that method is costly [4]. More recently, a flame spray pyrolysis (FSP) method was used to synthesize monoclinic Y2O3 particles with diameters up into the micrometer range. However, the polydisperse Y2O3 particles had mixed cubic and monoclinic phases. A critical particle diameter of approximately 1.5 µm was found. At the critical diameter, the probability was 50% for a particle to be either cubic or monoclinic. Particles significantly smaller than the critical diameter were all monoclinic, while those significantly larger were all cubic. [5]. To explore the potential applications for monoclinic Y2O3, one must first be able to synthesize phase-pure Y2O3 in sufficient quantities. This synthesis capability is also key to studying the interplay between surface energy and polymorphism [7]. The relationship between surface energy and polymorphism is a topic of profound importance in materials formation, especially on the nanometer scale [8]. This work was motivated by the above-mentioned reasons. The basic operating principle used in this study was that, if all the Y2O3 particles produced from the flame synthesis process were smaller than the critical size, then they would all have the monoclinic structure. Therefore phase-pure monoclinic Y2O3 particles may be generated via controlling the particle size. Herein we report the experimental methods for particle size control and the respective results. In particular, we describe the design and the successful use of a real impactor for achieving particle size control. To the
232
B. Guo et al. / Powder Technology 191 (2009) 231–234
Table 1 Summary of methods for synthesizing monoclinic Y2O3 particles Author
Method
Maximum process temperature (˚C)
Critical particle sizea
Yield
Hoekstra [1] Krauss et al. [2] Guo et al. [4]
High pressure diamond anvil Evaporation–condensation Flame aerosol process with gas-phase precursor Flame aerosol process with droplet precursor (flame spray pyrolysis)
~ 1000 ˚C 350 ˚C ~ 2700 ˚C
∞ (Bulk sample) ~ 10 nm Not observed (All particles monoclinic largest ~ 90 nm) ~ 1.5 µm (Mixed with larger Y2O3 particles that were cubic)
Unspecified ~ 150 mg/batch (inferred from sample description) 40–120 mg/h (batch mode; each run approximately 15 min) 200 mg /h (0.65 M precursor solution) and 10 mg/h (0.026M precursor solution)
Guo and Luo [5]
a
~ 2700 ˚C
Critical size is the size below which the Y2O3 particles had the monoclinic structure, and those larger had the cubic structure.
best of our knowledge, this is the first report of a real impactor being used in flame spray pyrolysis to obtain phase-pure product particles. 2. Experimental methods 2.1. Flame spray pyrolysis apparatus The flame spray pyrolysis apparatus is schematically shown in Fig. 1. It is similar to the apparatus used in a previous work [5], except that an optional impactor was incorporated in the apparatus in this study. The inner/outer diameters of the burner nozzle are 1.6 mm and 9.5 mm, respectively. The apparatus consisted of a 1.7 MHz atomizer, an atomization vessel, a furnace and a burner. The furnace and the burner have been described in detail elsewhere [4]. When the optional impactor was used, the precursor aerosol had to flow through the impactor first before entering the furnace and burner. The impactor removed the large droplets from the precursor aerosol. The precursor aerosol formed a self-sustained steady-state aerosol flame at the top of the burner. A co-flowing oxidant stream supported the flame. The furnace heated precursor aerosol to maintain a
sufficiently high flame temperature. The precursor aerosol underwent chemical reactions in the flame and became a Y2O3 aerosol. The postflame aerosol containing Y2O3 particles was drawn into a sampling tube by vacuum and the particles were collected on an alumina membrane filter (Whatman Inc., NJ). The synthesis apparatus was operated at atmospheric pressure. H2 was used as the fuel gas at a flow rate of 1 SLM (standard liter per minute). Pure O2 at 6 SLM was used as the oxidant stream to support the flame. The flame length was approximately 5 cm. The precursor solution was prepared by dissolving yttrium nitrate hexahydrate (chemical formula Y (NO3)3·6H2O, 99.9%, Alfa Aesar, Ward Hill, MA) in Nanopure® water (Barnstead, Dubuque, IO). The concentrations of the precursor solution used in this study were from 0.026 to 0.65 M. 2.2. Impactor design The objective of the impactor design was to ensure removal efficiency greater than 80% for 6-µm droplets. With a 0.65-M precursor solution, a 6-µm droplet would produce a 1.5-µm Y2O3 particle, assuming that one droplet becomes one final Y2O3 particle. Based on mass conservation one can readily infer the relation between a precursor droplet diameter and the resultant Y2O3 particle diameter, knowing the precursor concentration. This relation neglects the evaporation of Y2O3 particles and coagulation between Y2O3 particles. These assumptions will be discussed later. The primary dimensions of the impactor were determined using the relations given by Marple and Willeke [9]. The design gas was H2 at 1 SLM. The droplet density was 1120 kg/m3 (measured density for the 0.65 M precursor solution). The impactor had a round nozzle with an inner diameter of 2.87 mm. The nozzle inner diameter was selected by trial and error, so that the precursor droplets had the proper Stokes number, and hence desired removal efficiency in the impactor. The relation between removal efficiency and Stokes number (and the Reynolds number to a lesser degree) was found in the paper by Marple and Willeke [9]. The nozzleimpaction plate distance was 3.30 mm, selected based on the empirical relation for circular jet impactors [10]. Estimated removal (collection) efficiency and the corresponding Y2O3 particle diameter for several droplet sizes are given in Table 2. A schematic drawing of the impactor is shown in Fig. 2. 2.3. Computational fluid dynamics simulation for the impactor Computational fluid dynamic (CFD) simulation was used to verify the droplet removal efficiency for the impactor. The flow field through the impactor was simulated using FLUENT 6.2 (Fluent Inc., Lebanon, Table 2 Estimated removal efficiency and the corresponding Y2O3 particle diameter for several droplet sizes
Fig. 1. Schematic of flame spray pyrolysis apparatus for Y2O3 synthesis.
Droplet diameter, 0.65-M solution (µm) Stokes number Estimated efficiency of removal by impactor [9] Resultant Y2O3 particle diameter (µm)
4 0.20 b 20% 1.0
6 0.45 N 80% 1.5
6.8 0.58 N 90% 1.7
B. Guo et al. / Powder Technology 191 (2009) 231–234
233
This agrees with the expected performance of the impactor given in the Experimental Methods section. 3.2. Simulated particle removal efficiency Fig. 5 shows the particle removal efficiency calculated based on CFD simulation. It agrees well with the estimated efficiency shown in Table 2. The simulation shows impactor removal efficiency to be 100% for droplets larger than 7 µm. This is in agreement with the experimentally measured Y2O3 particle size distribution with impactor, in which particles larger than 1.5 µm are absent. 3.3. Crystal structure of Y2O3 particles
Fig. 2. A schematic drawing of the round jet impactor for removing large precursor droplets.
NH) with the appropriate models. The bulk fluid (continuous phase) flow field was solved using the Eulerian approach. After solving the flow field, the Discrete Phase Formulation was used to simulate particles trajectories based on the Lagrangian approach. The particle trajectories were calculated with the assumption that the particles neither influence the flow field nor each other. Trajectories for particle sizes in the range of 1–10 µm were generated. For each size, 500 trajectories were simulated with initial particle locations randomly placed at the inlet of the flow field. Particle removal efficiency for each size was then calculated based on the simulated particle trajectories. 3. Results
The XRD results in Fig. 6 show that with the impactor, the Y2O3 particles were phase-pure monoclinic. Without the impactor, the Y2O3 contained a significant amount of the cubic phase. This agrees with the fact that without the impactor a significant number of the Y2O3 particles were larger than the critical size, while the particles with the impactor were all smaller than the critical size. 4. Discussion 4.1. Size control via precursor concentration control As found in the previous study, when the precursor solution concentration was lowered to 0.026 M, the product Y2O3 particles were phase-pure monoclinic. Thus one could generate phase-pure monoclinic Y2O3 particles by lowering the precursor concentration [5]. However, that required the concentration to decrease by more than a factor of twenty. Because the particle yield is proportional to the precursor concentration, the low precursor concentration leads to
3.1. Effectiveness of impactor Typical TEM images of particles synthesized without the impactor and with the impactor are shown in Fig. 3. As can be seen in the TEM image, without the impactor, Y2O3 particles as large as 2 µm are present; while with the impactor, such large particles are absent. The effect of the impactor became evident when the particle size distribution was obtained by surveying at least 300 particles in the TEM images. The particle size distributions are shown in Fig. 4 for representative samples with and without the impactor. Without the impactor, a significant fraction of the particles are larger than 1.5 µm, the critical particle diameter. With the impactor, the largest particles are approximately 1.5 µm in diameter. Particles larger than 1.5 µm were essentially absent from the samples generated with the impactor. This suggests that the impactor had indeed removed precursor droplets larger than 6 µm (see Table 2 for size correlation).
Fig. 3. TEM images of Y2O3 particles synthesized (A) without and (B) with the impactor to remove large precursor droplets.
Fig. 4. Size distribution of Y2O3 synthesized without and with the impactor.
234
B. Guo et al. / Powder Technology 191 (2009) 231–234 Table 3 Effect of precursor concentration and use of impactor on synthesis yield
Fig. 5. Particle removal efficiency based on CFD simulation.
very low mass yield, as shown in Table 1. On the other hand, using the impactor, only the largest particles are stripped off the particle size distribution, hence a relatively high yield can be maintained while ensuring the particles to be phase-pure monoclinic. A comparison of the Y2O3 particle yield is given in Table 3. The flame used in this study was a laminar, quiet and steady flame. If particle disintegration and coagulation are negligible in the flame, then one precursor droplet becomes one product particle. In other flame spray pyrolysis processes, an air-blast type atomizer is integrated with the burner, and the precursor solvent is also a fuel [11]. The flame in those processes is a turbulent reacting jet, in which strong droplet/particle disintegration exists. One can find out that in such a process the onedroplet-one-particle relation does not apply [12].
Y(NO3)3·6H2O concentration (M)
0.65
Yield of Y2O3 particles (mg/h) Crystal structure
200 Mixed
0.65 (with impactor) 140 Phase-pure monoclinic
0.026 [5] 10 Phase-pure monoclinic
number flow rate may be calculated. Then the particle concentration in the flame was estimated to be 2.5 × 1012 1/m3, assuming it was the same as the precursor droplet number concentration in the fuel gas. Assuming the gas was air at 2700 K, for 800-nm particles, the estimated coagulation rate was about 1.5 × 1010 1/m3 s [10]. The particles’ residence time in the flame was in the order of 1 ms [4]. Within this time, coagulation between particles was negligible. In other words, vast majority of the Y2O3 particles would not experience significant coagulation due to the relatively low particle concentration and the short residence time in the flame. 5. Summary and conclusion By incorporating an impactor in a flame spray pyrolysis apparatus, the product Y2O3 particles were kept below 1.5 µm in diameter. The particles were all smaller than the critical diameter that determines the crystal structure of Y2O3 in this flame synthesis process. As a result, the Y2O3 particles were phase pure with the monoclinic crystal structure. The result from this study shows that, in spray pyrolysis where a particle size effect on crystal structure exists, one can use a simple impactor to obtain phasepure particles of one crystal structure. Acknowledgement
4.2. Negligible Y2O3 particle coagulation In this study it was assumed that each precursor droplet became one Y2O3 particle. Then we assumed that by removing those large droplets that would make Y2O3 particles larger than the 1.5 µm, we could prevent the formation of cubic Y2O3 particles. However, in an aerosol smaller particles coagulate to form larger particles. Thus particles larger than 1.5 µm may be formed from smaller particles. Therefore, the coagulation rate was estimated for the Y2O3 particles in the flame. Based on the precursor solution atomization rate (5 mL/ min) and an estimated average droplet diameter of 4 µm, the particle
Fig. 6. XRD of Y2O3 synthesized without and with the impactor; filled diamonds mark major distinctive peaks of cubic Y2O3.
Financial support for this work was provided by Texas Engineering Experiment Station and Texas A&M University. The Microscopy and Imaging Center at Texas A&M University made available the electron microscope used in this study.
References [1] H.R. Hoekstra, Phase relationships in the rare earth sesquioxides at high pressure, Inorganic Chemistry 5 (5) (May 1966) 754–757. [2] W. Krauss, R. Birringer, Metastable phases synthesized by inert-gas-condensation, Nanostructured Materials 9 (1–8) (1997) 109–112. [3] X. Qin, Y.G. Ju, S. Bernhard, N. Yao, Flame synthesis of Y2O3: Eu nanophosphors using ethanol as precursor solvents, Journal of Materials Research 20 (11) (Nov 2005) 2960–2968. [4] B. Guo, A. Harvey, S.H. Risbud, I.M. Kennedy, The formation of cubic and monoclinic Y2O3 nanoparticles in a gas-phase flame process, Philosophical Magazine Letters 86 (7) (Jul 2006) 457–467. [5] B. Guo, Z.P. Luo, Particle size effect on the crystal structure of Y2O3 particles formed in a flame aerosol process, Journal of the American Ceramic Society 91 (5) (2008) 1653–1658. [6] B. Guo, A.S. Harvey, J. Neil, I.M. Kennedy, A. Navrotsky, S.H. Risbud, Atmospheric pressure synthesis of heavy rare earth sesquioxides nanoparticles of the uncommon monoclinic phase, Journal of the American Ceramic Society 90 (11) (2007) 3683–3686. [7] P. Zhang, A. Navrotsky, B. Guo, I. Kennedy, A.N. Clark, C. Lesher, et al., Energetics of cubic and monoclinic yttrium oxide polymorphs: phase transitions, surface enthalpies, and stability at the nanoscale, Journal of Physical Chemistry C 112 (2008) 932–938. [8] A. Navrotsky, Energetics of nanoparticle oxides: interplay between surface energy and polymorphism, Geochemical Transactions 4 (Nov 18 2003) 34–37. [9] V.A. Marple, K. Willeke, Impactor design, Atmospheric Environment 10 (10) (1976) 891–896. [10] W.C. Hinds, Aerosol Technology: Properties, Behavior, and Measurement of Airborne Particles, 2nd edWiley, New York, 1999. [11] L. Madler, H.K. Kammler, R. Mueller, S.E. Pratsinis, Controlled synthesis of nanostructured particles by flame spray pyrolysis, J Aerosol Sci 33 (2) (Feb 2002) 369–389. [12] D. Dosev, B. Guo, I.M. Kennedy, Photoluminescence of Eu3+: Y2O3 as an indication of crystal structure and particle size in nanoparticles synthesized by flame spray pyrolysis, J Aerosol Sci 37 (3) (Mar 2006) 402–412.