(565b) Dielectrophoresis in Porous Media: Electrically Switchable Filtration

Dielectrophoresis in Porous Media: Electrically Switchable Filtration

Georg R. Pesch, Fei Du, Jorg Thöming, Michael Baune

Chemical Engineering -- Recovery and Recycling (VdW), Department of Production Engineering and Center for Environmental Research and Sustainable Technology, University of Bremen

 

 

Dielectrophoresis (DEP) describes the movement of charged or uncharged matter in inhomogeneous electric fields and is a versatile and promising technique for particle manipulation. It is caused by the accumulation of polarization charges at material interfaces between the particles and the surrounding medium. The DEP force acting on particles is directly related to the electric field strength and its spatial change and depends mainly on the particle volume. Hence, the movement of very small particles (i.e., nanoparticles) requires vast electric field gradients. These gradients could be achieved by either using technically mature electrode configurations or by letting an originally homogeneous electric field scatter at solid obstacles, as it is the case in insulator-based (electrodeless) DEP (iDEP).

In the presence of an excitatory field, an obstacle of different material than the liquid medium will polarize. This effect is caused by charge separation induced by the excitatory field in combination with the different dielectric properties of the liquid medium and the obstacle's material. The resultant polarization charges itself will cause a highly inhomogeneous electric field, which is very suitable for targeted particle motion.

We applied this technique in the concept of dielectrophoretically driven filtration (cf. Fig 1) [1]: In this, a conventional deep-bed filtration process is altered so that the filter material is embedded between two electrodes. Because the pores of the filter are much larger than the target particle diameter, no noteworthy mechanical retention of particles occurs in the absence of an electric field. When applying an ac voltage on the electrodes, however, the generated electric field scatters at the porous dielectric. This results in a highly inhomogeneous field inside of the pores with maximum electric field values at their surface. These field inhomogeneities give rise to a DEP force acting on particles passing through the pores. It will direct them towards the pore surface, where the particles are trapped and reside until the electric field is turned off. Hence, we can separate particles in the nm-size range without filter cake formation by using filters with pore sizes being several orders of magnitude larger than the particle diameter. This effectively reduces pressure loss in comparison to conventional filtration processes and gives the possibility to easily turn off the filtration effect by switching off the electric field, enabling safe and easy resuspension of trapped NPs. In semi-continuous operation mode we reached separation and recovery efficiencies for dp = 340 nm particles in dpore ∼ 100 μm pores of up to 65% with a flow rate of 90 L h-1 m-2, as a proof of principle.

In this first experiment, module design and choice of filter material was not based on deep insight into dielectrophoretic effects in porous structures. In order to better understand the trapping and scattering mechanisms inside the pores we abstracted the originally complex 3 dimensional porous geometry to a quasi 2 dimensional flow channel containing an array of post-like field obstacles; a geometry which very much resembles insulator-based DEP flow channels. We developed a novel method to analyze and evaluate the polarization of such obstacles. It is based on the theory that the polarization of quasi 2-d obstacles in homogeneous electric fields can be expressed by superposition of an infinite amount of multipoles. The multipole coefficients can be extracted by any solution for the electrostatic potential by numerical integration. This method allows easy reproduction of the polarization field after the coefficients have been extracted. The possibility to compare the coefficients of different structures yields an easy way to benchmark the obstacles for DEP particle trapping.

For the first time, a thorough analysis on the influence of the obstacles cross-sectional geometry and material on their DEP particle trapping efficiency is conducted. From the polarization field obtained by the multipole coefficients we derive an equation for the DEP force exerted on spherical particles (based on the point-dipole approximation). With this equation and a superimposed fluid flow we are able to effectively describe effectiveness of a given obstacle for attracting and trapping particles out of a fluid stream using positive DEP (cf. Figs. 2 and 3). With this method we are able to derive crucial parameters for the dielectrophoretical filtration process, such as ideal flow rate, pore radius, porosity, pore geometry, and electric field strength for a given particle diameter. Additionally, this enables us to predict the separation efficiency and transfer function of a given filter-particle--system.

 

[1] G. R. Pesch, F. Du, U. Schwientek, C. Gehrmeyer, A. Maurer, J. Thöming, M. Baune: Recovery of submicron particles using high-throughput dielectrophoretically switchable filtration. Sep. Purif. Technol. 132 (2014) 728--735.

 

 

Figure 1: a) and b) show a sketch of the DEP filtration system. a) No voltage is applied and the particles simply flow through the comparably large pores. b) When sufficient voltage is applied, a DEP force is acting on the particles and directs them towards the surface oft he pores. C) Particle separation efficiency as a function of filter pore size and filter thickness at a specific flowrate of q = 40 L h-1 m-2.

 

Figure 2: Post-like obstacle (as seen from top). Electric field applied perpendicular to the longitudinal axis. Particles are passing by the obstacle in distances H and L and are carried by a stationary fluid with velocity v0.

Figure 3: Critical particle diameter dp,crit which could be separated from a stationary fluid flow by a polarized obstacle with circular cross-section for different particle-to-obstacle distances H and different ratios between the obstacle's permittivity and the surrounding liquid medium's permittivity (cf. Fig 2).