Experiments and modeling on the deacidification of agglomerates of nanoparticles in a fluidized bed☆
Graphical abstract
The deacidification of the fumed silica AEROSIL® 200 was studied experimentally in a batch fluidized bed in the temperature range from 250 °C to 400 °C. A mathematical reactor model was developed for the adsorption and for the surface reaction on these highly agglomerated nanoparticles
Introduction
Fumed oxides (SiO2, Al2O3, TiO2) are industrially produced in flame reactors. In the case of fumed silica, a chlorosilane vapor (e.g. SiCl4) is mixed with air and hydrogen and hydrolysis takes place in a flame at temperatures well above 1000 °C. A simplified process schematic is given in Fig. 1. Nano-sized primary particles are formed, which sinter together to aggregates of sub micron size. The morphology of these aggregates is determined by the reaction conditions in the flame reactor. Subsequently, these aggregates adhere to each other to form three dimensional agglomerates, which can be in the several hundred micron range in size. Fig. 2 shows SEM pictures of AEROSIL® 200 at two magnifications showing the aggregate and agglomerate structure of fumed silica. The overall porosity of these agglomerates can exceed 98% and a fixed bed of such material has bulk densities of approximately 20 g/l only. Due to the small primary particles, industrially produced fumed oxides have large specific surface areas of up to 400 m2/g. On their large specific surface there are functional chemical groups, in case of fumed silica these are silanol (Si–OH) and siloxane (Si–O–Si) groups. Some of these functional groups (e.g. the silanol groups) can readily interact with other chemicals or they can be modified in the production process itself.
This unique morphology paired with a wide array of possible surface chemistry make fumed oxides a necessary compound in many products. Fumed silica for example is used in the silicone industry to provide the desired rheology and mechanical strength in silicone adhesives and silicone rubbers. Fumed alumina is used for example to treat ink jet paper for improved ink absorbance and fumed titania is used in cosmetic applications such as sun protection. These same properties also require special considerations in the processing of these solids which often go beyond the conventional text book knowledge for solids processing.
In the industrial processes for fumed oxide manufacturing, fluidized beds are used to remove the byproduct HCl from the fumed oxides (so called deacidification, see Fig. 1) or for chemical modification of the functions surface groups (e.g. to make a hydrophilic fumed silica hydrophobic). Whereas research on gas/solids fluidized bed is extensive, the fluidization of fumed oxides has only been studied recently [1], [2], [3]. Some fumed oxides can be fluidized well (e.g. fumed silica with large specific surface area) and the fluidization is characterized by large bed expansion and no channeling. Whether bubbles are formed, depends on certain fluidization parameters, such as gas distributor type or fluidization velocity [4]. This type of fluidization of nanoparticles has been classified as agglomerate particulate fluidization (APF) [1]. On the other hand, some fumed oxides (e.g. fumed titania or fumed silica with low surface area) are difficult to fluidize and behave like Geldart B type solids with little bed expansion, channeling, and if fluidization can be initiated, it shows strong bubbling. This type of fluidization behavior has been named agglomerate bubbling fluidization or ABF. The research group of Prof. Pfeffer at NJIT (e.g. Yu et al. [5]) has investigated methods to assist fluidization of such particles.
Although fluidized beds have been used for decades in the production of fumed oxides, their design and operation are more based on experience, rather than on a fundamental understanding of the physico-chemical principles governing this process step. No reactor model has been reported so far, which can describe adsorption and chemical reaction of such low bulk density particulate systems in fluidized beds. The presentation of such a model and the comparison with experimental data for the deacidification of the fumed silica AEROSIL® 200 is the scope of this work.
Section snippets
Experimental
The deacidification experiments were carried out in a batch fluidized bed reactor. Fig. 3 shows a P&I Diagram of the reactor with the conditioning of the fluidizing gas and the sampling and discharge of the particles. The reactor is a cylindrical glass column (Borosil Glass, QVC Co., Germany) with an overall height of 2050 mm and an inner diameter of 210 mm. A 5 mm thick porous sintered metal plate (Cr–Ni Steel 1.4404, Tridelta Siperm Co., Germany) with 20 μm pores serves as the gas distributor
Model
Fig. 5 shows a schematic of the reactor model for the batch fluidized bed employed in the experiments. The extended fluidized bed is divided in three phases. The first phase is the fluidized agglomerates, which are assumed to be spheres with constant diameter dF and a porosity ɛF. These agglomerates are suspended in the suspension phase exhibiting the porosity ɛg. Solid free spherical bubbles of constant size make the third phase.
The fumed oxide agglomerates are fluidized with gas of the
Results and discussion
The objective of this work is to identify the process parameters which influence the rate of deacidification of fumed silica and to quantify these parameters for process optimization and design. First a standard deacidification experiment was defined to serve as a baseline for variations of the parameters temperature, moisture content in fluidizing gas, fluidization velocity, and mass of bed. The experimental data presented in the following are compared to the modeling results and used in parts
Conclusion
The deacidification of the fumed silica AEROSIL® 200 was studied experimentally in a batch fluidized bed in the temperature range from 250 °C to 400 °C. For a well fluidized bed, the temperature and the steam concentration in the fluidizing gas are the determining parameters for the overall rate of deacidification. If the bed is not well fluidized, e.g. because it is too shallow, or it is fluidized near the point of minimum fluidization velocity, the rate of deacidification drops because
Symbols and Indices
AFB m2 Cross section of the fluidized bed a m2 Surface of a sphere (bubble or agglomerate) kg/kg Mass fraction of component i in the SiO2 particle kg/kg Mass fraction of Si–Cl which cannot be removed at a given temperature c kg/m3 Mass concentration in a given phase d m Diameter EA kJ/mol Activation energy H m Height of the expanded bed H0 m Height of fixed bed Kp 1/mbar Equilibrium constant for adsorption Kc m3/kg Equilibrium constant for adsorption k0 kg/s Pre factor in Arrhenius equation kR kg/s Rate constant for
Acknowledgements
J. Flesch would like to thank Professor Robert Pfeffer and his former PhD student Dr. Jose Quevedo for many fruitful discussions on the fluidization of fumed silica in the frame of a research cooperation (not this work) between NJIT and Evonik Degussa. The work presented here was part of a research cooperation between the Universität Karlsruhe and Evonik Degussa GmbH (both in Germany). J. Flesch would like to thank his colleagues Dr. Kerner and Prof. Riemenschneider for their support and Prof.
References (12)
- et al.
Aggregation behavior of nanoparticles in fluidized beds
Powder Technol.
(2005) - et al.
Effect of bubble interaction on interphase mass transfer in gas fluidized beds
Chem. Eng. Sci.
(1981) - et al.
Fluidization and agglomerate structure of SiO2 nanoparticles
Powder Technol.
(2002) - et al.
Gas fluidization characteristics of nanoparticle agglomerates
AIChE J.
(2005) - J. Flesch, Untersuchungen zu den Gleichgewichten und zur Reaktionskinetik bei der Entsäuerung pyrogener Kieselsäuren in...
- et al.
Enhanced fluidization of nanoparticles in an oscillating magnetic field
AIChE J.
(2005)
Cited by (0)
- ☆
Dedicated to Professor Robert Pfeffer on the occasion of his 70th birthday.