Characteristics of non-magnetic nanoparticles in magnetically fluidized bed by adding coarse magnets
来源期刊:中南大学学报(英文版)2011年第5期
论文作者:周立 刁润丽 周涛 Hiroyuki Kage Yoshihide Mawatari
文章页码:1383 - 1388
Key words:non-magnetic nanoparticles; magnetic fluidization; agglomerate; coarse magnet
Abstract:
The fluidization behavior of SiO2, ZnO and TiO2 non-magnetic nanoparticles was investigated in a magnetically fluidized bed (MFB) by adding coarse magnets. The effects of both the amount of coarse magnets and the magnetic field intensity on the fluidization quality of these nanoparticles were investigated. The results show that the coarse magnets added to the bed lead to a reduction in the size of the aggregates formed naturally by the primary nanoparticles. As the macroscopic performances of improved fluidization quality, the bed expansion ratio increases whilst the minimum fluidization velocity decreases with increasing the magnetic field intensity, but for TiO2 nanoparticles there exists a suitable magnetic field intensity of 0.059 6 T. The optimal amounts of coarse magnets for SiO2, ZnO and TiO2 non-magnetic nanoparticles are 40%, 50% and 60% (mass fraction), respectively. The bed expansion results analyzed by the Richardson-Zaki scaling law show that the exponents depend on both the amount of coarse magnets and the magnetic field intensity.
J. Cent. South Univ. Technol. (2011) 18: 1383-1388
DOI: 10.1007/s11771-011-0850-z
ZHOU Li(周立)1, 2, DIAO Run-li(刁润丽)1, ZHOU Tao(周涛)1, Hiroyuki Kage3, Yoshihide Mawatari3
1. School of Chemistry and Chemical Engineering, Central South University, Changsha 410083, China;
2. Department of Mechanical Engineering, Hunan Institute of Technology, Hengyang 421002, China;
3. Department of Applied Chemistry, Kyushu Institute of Technology, Tobata, Kitakyushu, Japan
? Central South University Press and Springer-Verlag Berlin Heidelberg 2011
Abstract: The fluidization behavior of SiO2, ZnO and TiO2 non-magnetic nanoparticles was investigated in a magnetically fluidized bed (MFB) by adding coarse magnets. The effects of both the amount of coarse magnets and the magnetic field intensity on the fluidization quality of these nanoparticles were investigated. The results show that the coarse magnets added to the bed lead to a reduction in the size of the aggregates formed naturally by the primary nanoparticles. As the macroscopic performances of improved fluidization quality, the bed expansion ratio increases whilst the minimum fluidization velocity decreases with increasing the magnetic field intensity, but for TiO2 nanoparticles there exists a suitable magnetic field intensity of 0.059 6 T. The optimal amounts of coarse magnets for SiO2, ZnO and TiO2 non-magnetic nanoparticles are 40%, 50% and 60% (mass fraction), respectively. The bed expansion results analyzed by the Richardson-Zaki scaling law show that the exponents depend on both the amount of coarse magnets and the magnetic field intensity.
Key words: non-magnetic nanoparticles; magnetic fluidization; agglomerate; coarse magnet
1 Introduction
Nanoparticles have attracted a great deal of attentions because of their increasing potential in wide industrial application. Due to their unique properties, such as very small primary particle size and very large surface area per unit mass, nano-structured materials are being used in the manufacture of drugs, cosmetics, foods, plastics, catalysts, and energetic and bio-materials. Therefore, it is necessary to develop processing technologies which can handle large quantities of nano-sized particles, for example, mixing, transporting, modifying the surface properties (coating) and downstream processing of nanoparticles to form nano-composites. Before processing of nano-structured materials, the nano-sized particles have to be well dispersed. However, nanoparticles often cause undesired agglomeration accounting for the strong cohesive forces, leading to handling troubles and reducing productivity. Gas-solid fluidization is one of the best techniques available to disperse nanoparticles effectively [1-7]. Addition of coarse particles to a bed of nanoparticles could break the aggregates into small ones even though they will be formed again. Therefore, the choice of an external field includes matter of sources available in laboratory: magnetic, vibrations, acoustic field, etc, all of which can act on the particles suspended in the fluidizing flow.
Magnetically assisted fluidized beds share many of the advantages of both packed beds and conventional fluidized beds. In magnetically assisted (or magnetically driven) fluidization, the particles aggregate into agglomerates which reduce the fluid-solid contact area. The agglomerate break-up creates a renewable fluid-solid area by continuous formation and destruction and together the high fluid slip velocity ensures effective heat and mass transfer operations [8]. In contrast to the coarse magnetic beds, nanoparticles are fluidized with relatively low gas velocities and the drag forces are not enough to destroy the particle agglomerates. Besides, most of nanoparticles are not magnetic in nature. Thus, the optimal solution is to add a certain amount of coarse magnets in the bed.
The fluidization of cohesive non-magnetic particles has been investigated by many researchers [9-13]. In their researches, the particles used were micro-size particles or nanoparticles of low density (~40 kg/m3) prepared by CVD method. ZENG et al [14-15] studied the behavior of mixtures of non-magnetic nanoparticles in a magnetically assisted fluidized bed. HRISTOV [16] reviewed the hydrodynamics of all systems in the previous parts of the series (G-S, L-S and G-L-S) as a foundation assuring proper understanding of their mass transfer performances in magnetic field assisted fluidization. However, the behavior of single-component nanoparticles prepared by liquid phase method in a magnetically fluidized bed has not been reported so far.
In the present work, SiO2, ZnO and TiO2 non- magnetic nanoparticles prepared by liquid phase method are fluidized by adding coarse magnets to a magnetic bed. The bed expansion and pressure drop in the bed of coarse magnets and non-magnetic nanoparticles were measured. The Richardson-Zaki equation was used to analyze the bed expansion of single-component non-magnetic nanoparticles.
2 Experimental
The experiments were performed in a glass column (50 mm in diameter and 1 000 mm in height) with a sintered gas distributor. Dried compressed air was used as fluidizing agent at atmospheric pressure and room temperature. The superficial gas velocity and pressure drop were measured by the rotameter and U-type manometer, respectively (see Fig.1).
Fig.1 Schematic diagram of experimental set-up: 1—Fluidized bed; 2— Ruler; 3—Electromagnetic coil; 4—Rotameter; 5—Silica gel drying column; 6—Compressed air; 7—U-type manometer; 8—DC current
Uniform axial (parallel to the column axis and fluid flow) steady-state magnetic field created by four electromagnetic coils (80 mm in inner diameter, 200 mm in outer diameter and 35 mm in height) was used. The field intensity was varied by changing the current strength of a DC electrical supply from 0 to 6 A and was calculated by the expression:
H=4?0NI/(2R) (1)
where H is the magnetic field intensity, and the operation range is 0-0.1 T; ?0 is a constant (12.56×10-7 H/m); N is the circle of coils; I is current (A); R is the radius of coils (m).
Steel particles (dp=2 mm and density of 7 780 kg/m3) were used as coarse magnets and the adding amount changes in the range of 0-60% (mass fraction). The properties of nanoparticles used in the experiments are listed in Table 1. At the beginning of each experiment, some amounts of coarse magnets were added in the bed firstly forming a magnetic layer at the gas distributor, then the SiO2, ZnO and TiO2 nanoparticles were added forming the bed undergoing fluidization. All experiments were repeated four times and it was found out that no identical curves could be obtained especially at lower gas velocities because the pressure drop was fluctuate due to channeling or plugs.
Table 1 Properties of nanoparticles used in experiments
3 Results and discussion
3.1 Behavior of nanoparticles in traditional fluidized bed
As discussed earlier [14-15], SiO2, ZnO and TiO2 nanoparticles can not be fluidized normally with bubbles. Commonly, the bed breaks into plugs lifted by the gas. The plugs are protruded by cracks and voids causing a gas bypass. Such non-uniformities result in violent pulsation of the pressure drop across the bed and uncontrollable bed expansion. Increasing the gas velocity leads to the break-up of the plugs, enlargement in the channel size and formation of a bed of fluidized particle agglomerates rather than single particles. Under such conditions, the bed expands slightly with a top surface perturbed by bursting agglomerates and elutriated particles. Additionally, segregation by size occurs along the bed depth: large particle agglomerates exist near the distributor whilst near the bed top the agglomerates are smaller in size (see Figs.2 and 3).
Fig.2 Pressure drop curves of nanoparticles in traditional fluidized bed
Fig.3 Bed expansion ratio of nanoparticles in traditional fluidized bed
3.2 Fluidization of non-magnetic nanoparticles in MFB
It is worth to pay an attention that improvement of the fluidization behavior is attained by adding some amount of coarse magnets as a layer just above the gas distributor. Their motions lead to oscillating gas jets entering the nanoparticle bed above. In these ways, the formation of permanent channels through the fine bed above is avoided. The bed expands, as the gas flow increases, in a manner quite smoother than in absence of the bottom magnetic layer. Figures 4-9 show the fluidization behavior of non-magnetic nanoparticles in MFB. It can be seen that SiO2, ZnO and TiO2 nanoparticles can be fluidized normally in MFB only in a specific adding amount of coarse magnets or magnetic field intensity.
3.3 Effects of amount of coarse magnets on behavior of nanoparticles
The pressure drop curves of SiO2 nanoparticles are shown in Fig.4 when the magnetic field intensity H= 0.059 6 T and the amounts of coarse magnets are 20%, 30%, 40% and 50% (mass fraction), respectively. The fluidization quality is rarely improved when the amount of coarse magnets is 20% or 30%. Plugs and channels occur in turn with increasing the gas velocity. But the pressure drop is not stable and elutriation occurs at higher gas velocities. For the amount of coarse magnets of 40%, the bed expansion ratio and pressure drop are increased continuously and the pressure drop reaches steady state (fluidization) when the gas velocity is 0.050 81 m/s. However, the pressure drop curve of the amount of coarse magnets of 50% is similar to that of 30%. This is because if the amount of coarse magnets is low, their motions in magnetic field cannot break up all the big agglomerates. Therefore, the fluidization quality is rarely improved. If the percentage of coarse magnets is excessively large, coarse magnets may hinder the fluidization of nanoparticles. So, there exists a suitable amount of coarse magnets of 40%.
Fig.4 Pressure drop of SiO2 nanoparticles at H=0.059 6 T with different amounts of coarse magnets
Fig.5 Pressure drop of ZnO nanoparticles at H=0.059 6 T with different amounts of coarse magnets
Figure 5 illustrates the pressure drop curves of ZnO nanoparticles when magnetic field intensity H=0.059 6 T and the amounts of coarse magnets are 30%, 40%, 50% and 60% (mass fraction), respectively. For the amount of coarse magnets of 30%, there are large voids at the edge of the bed. For the amount of coarse magnets of 40%, the height of fixed bed in the bottom is decreased evidently and the fluidization quality is slightly improved. With the amount of coarse magnets increasing, the agglomerates are broken up by the magnetic chain formed by coarse magnets. For the amounts of coarse magnets of 50% and 60%, the homogeneous fluidization is achieved.
The pressure drop curves of TiO2 nanoparticles are presented in Fig.6 when magnetic field intensity H= 0.047 7 T and the amounts of coarse magnets are 30%, 40%, 50% and 60%, respectively. For the amount of coarse magnets of 30%, the bed is not fluidized. With the amount of coarse magnets increasing, the fluidization quality is slightly improved. For the amount of coarse magnets of 60%, the bed is well fluidized at the gas velocity of but the loss of elutriation increases at higher gas velocities.
Fig.6 Pressure drop of TiO2 nanoparticles at H=0.047 7 T with different amounts of coarse magnets
Fig.7 Pressure drop of SiO2 nanoparticles with 50% (mass fraction) coarse magnets
Fig.8 Pressure drop of ZnO nanoparticles with 50% (mass fraction) coarse magnets
Fig.9 Pressure drop of TiO2 nanoparticles with 50% (mass fraction) coarse magnets
3.4 Effects of magnetic field intensity on behavior of nanoparticles
The pressure drop curves of non-magnetic nanoparticles in MFB are shown in Figs.7-9 at the same amount of coarse magnets and different magnetic field intensities. It can be seen that the fluidization quality of SiO2 and ZnO nanoparticles can be improved with increasing the magnetic field intensity. This means that the minimum fluidization velocity is a function of the magnetic field intensity. But for TiO2 nanoparticles, there exists a suitable magnetic field intensity. The experimental results of the bed expansion are listed in Table 2. The bed expansion ratios of SiO2 and ZnO nanoparticles increase with increasing the magnetic field intensity, whilst for TiO2 nanoparticles there is little change in the bed expansion ratio. However, coarse magnets may gather together when the magnetic field intensity is excessively high and they could not break agglomerates and bubbles anymore.
Table 2 Maximum bed expansion ratios for nanoparticles with different magnetic field intensities
3.5 Applicability of Richardson-Zaki equation
The bed expansion profiles corresponding to the stable (homogenous) range of fluidization are commonly modeled by the Richardson-Zaki (R-Z) scaling law (Eq.(2)) originally developed for liquid-solid systems with negligible interparticle forces [17-18]. But, it is also fit for gas-solid systems [5, 10, 19-21]. For instance, nanoparticles with primary particle size of 7-500 nm can form agglomerates, and the bed expansion obeys the R-Z equation:
(2)
where Ug is the superficial gas velocity (or gas velocity); U t is the terminal velocity; e is the bed voidage; n is the R-Z exponent and it is determined by the particle properties and bed properties. Generally, the larger the value of n, the better the fluidization quality.
For a nanoparticle bed, the density of the agglomerates is hypothesized to remain almost constant before and during the fluidization. So, the voidage ε of the bed formed by particle aggregates can be expressed by mass conservation law:
(3)
where e0 is the initial bed voidage, h0 is the initial height of bed, and h is the height of fluidized bed. The range of e0 is 0.2-0.25 for different primary size (7-20 nm) and lower bulk density SiO2 nanoparticles [5]. So, the value of e0 for SiO2 in the experiment is chosen as 0.23. Primary density and bulk density are large and primary particle size is small for ZnO and TiO2 particles, so the value of e0 is chosen as 0.18 and 0.16, respectively, smaller than that of SiO2 particles.
By combining Eqs.(2) and (3), the relation between the superficial gas velocity and the agglomerate terminal velocity in log–log coordinates can be written as
(4)
Through the above discussion, the values of Ug and e are known, so the linear fit can be drawn in lg Ug-lg ε coordinates, as shown in Fig.10. From linear regression, the slope n can be obtained and listed in Table 3.
Table 3 Values of n under different mass fraction of coarse magnets
It can be seen from Table 3 that nanoparticle beds undergoing fluidization exhibits almost linear bed expansion profiles in lg Ug-lg ε coordinates. Under the same range of superficial gas velocities, the values of n increase with increasing the amount of coarse magnets. This might be attributed to the increasing agglomerate broken by magnets. It is also clear that the exponent n of SiO2 nanoparticles is relatively higher compared with ZnO and TiO2 nanoparticles, indicating that the quality of SiO2 is better than ZnO and TiO2 in MFB. But the fluidization behavior of SiO2, ZnO and TiO2 nanoparticles is all improved in MFB compared with in traditionally fluidized bed. It could be attributed to the effect of the motion of coarse magnets.
Fig.10 Bed voidage versus superficial gas velocity for nanoparticles under different mass ratio of coarse magnets: (a) SiO2; (b) ZnO; (c) TiO2
4 Conclusions
1) The fluidization quality of SiO2 and ZnO nano- particles can be improved significantly with increasing the magnetic field intensity, but for TiO2 nanoparticles there exists a suitable magnetic field intensity of 0.059 6 T.
2) SiO2, ZnO and TiO2 non-magnetic nanoparticles can be fluidized at the amount of coarse magnets of 40%, 50% and 60% (mass fraction), respectively.
3) The calculated exponents of Richardson-Zaki equation depend on both the amount of coarse magnets and intensity of the field applied.
References
[1] MATSUDA S, HATANO H, MURAMOYO T, TSUTSUMI A. Modeling for size reduction of agglomerates in nanoparticle fluidization [J]. AIChE J, 2004, 50(11): 2763-2771.
[2] ZHOU T, LI H Z, SHINOHARA K. Agglomerating fluidization of group C particles: Major factors of coalescence and breakup of agglomerates [J]. Advanced Powder Technol, 2006, 17(2): 159-166.
[3] ZHOU T, LI H Z. Estimation of agglomerate size of cohesive particles during fluidization [J]. Powder Technol, 1999, 101(1): 57-62.
[4] YU Q, DAVE R, ZHU C. Enhanced fluidization of nanoparticles in an oscillating magnetic field [J]. AIChE J, 2005, 51(7): 1971-1979.
[5] ZHU C, YU Q, DAVE R. Gas fluidization characteristics of nanoparticle agglomerates [J]. AIChE J, 2005, 51(2): 426-439.
[6] HAKIM L, PORTMAN J, CSAPER M, WEIMER A. Aggregation behavior of nanoparticles in fluidized beds [J]. Powder Technol, 2005, 160(3): 149-160.
[7] LI H Z, TONG H. Multi-scale fluidization of ultrafine powders in a fast-bed-riser/conical-dipleg CFB loop [J]. Chem Eng Sci, 2004, 59(8/9): 1897-1904.
[8] HRISTOV J Y. Magnetic field assisted fluidization—A unified approach: Part 5. A hydrodynamic treatise on liquid-solid fluidized beds [J]. Review of Chem Eng, 2006, 22(4/5): 195-377.
[9] WANG X S, RAHMAN F, RHODES M J. Nanoparticle fluidization and Geldart’s classification [J]. Chem Eng Sci, 2007, 62(13): 3455- 3461.
[10] ZHU Q S, LI H Z. Study on magnetic fluidization of group C powders [J]. Powder Technol, 1996, 86(2): 179-185.
[11] HRISTOV J Y. Simple bed expansion correlations for magnetically assisted gas-fluidized tapered beds [J]. International Review of Chemical Engineering-Rapid Communications, 2009, 1(4): 316-323.
[12] YANG J, SLIVA A, BANERJEE A, DAVE R N, PFEFFER R. Dry particle coating for improving the flowability of cohesive powders [J]. Powder Technol, 2005, 158(1/2/3): 21-33.
[13] LU X S, LI H Z. Fluidization of CaCO3 and Fe2O3 particle mixtures in a transverse rotating magnetic field [J]. Powder Technol, 2000, 107(1/2): 66-78.
[14] ZENG P, ZHOU T, CHEN G Q, ZHU Q S. Behavior of mixed ZnO and SiO2 nano-particles in magnetic field assisted fluidization [J]. China Particuology, 2007(5): 169-173.
[15] ZENG P, ZHOU T, YANG J S. Behavior of mixtures of nano-particles in magnetically assisted fluidized bed [J]. Chem Eng and Process, 2008, 47(1): 101-108.
[16] HRISTOV J Y. Magnetic field assisted fluidization—A unified approach: Part 7. Mass Transfer: Chemical reactors, basic studies and practical implementations thereof [J]. Reviews in Chemical Engineering, 2009, 25(1/2/3): 1-254.
[17] RICHARDSON J F, ZAKI W N. Sedimentation and fluidization: Part I [J]. Transactions of the Institution of Chemical Engineers, 1954, 32: 35-54.
[18] CHEN G Q, ZHOU T, LI H Z. Characteristics of anthraquinone hydrogenation catalysts in a liquid-solid fluidized bed [J]. The Canadian J of Chem Eng, 2008, 86: 288-292.
[19] YANG J S, ZHOU T, SONG L Y. Agglomerating vibro-fluidization behavior of nano-particles [J]. Advanced Powder Technol, 2009, 20(2): 158-163.
[20] WANG Y, GU G, WEI F, WU J. Fluidization and agglomerate structure of SiO2 nanoparticles [J]. Powder Technol, 2002, 124(1/2): 152-159.
[21] VALVERDE J M, CASTELLANOS A. Fluidization of nanoparticles: A modified Richardson–Zaki law [J]. AIChE J, 2006, 52(2): 838-842.
(Edited by HE Yun-bin)
Foundation item: Project(20776163) supported by the National Natural Science Foundation of China; Project(20070533121) supported by the PhD Programs Foundation of Ministry of Education of China; Project supported by the NSFC-JSPS Cooperation Program
Received date: 2010-09-06; Accepted date: 2011-01-19
Corresponding author: ZHOU Tao, Professor; Tel: +86-731-88876605; E-mail: zhoutao@mail.csu.edu.cn