Highly efficient removal of ammonia nitrogen from wastewater by dielectrophoresis-enhanced adsorption

Effects of processing factors on the removal efficiency

The factors of removal of NH₃-N by DEP-enhanced adsorption are given in Figs. 2–6.

Characterization of the zeolite particles

To find out how adsorption and adsorption-DEP processes affected the zeolite particles, we performed SEM examination on the samples taken after each process (Fig. 7).

Scanning electron microscopy image taken from the sample after an adsorption-only process did not show any morphological change in the particles (Fig. 7A). However, the particles were broken down by the adsorption-DEP process, which indicates the important effect of DEP on the particles (Fig. 7B). We have also examined the electrodes after the adsorption-DEP by SEM, and the result is shown in Figs. 7C and 7D. It can be seen that both positive and negative electrodes were deposited by zeolite particles.

Theory of DEP

We would like to propose the following mechanism to explain the DEP process: a non-uniform electric field generated in the DEP device polarizes adsorbents and induces a dipole moment on each of them. As a consequence, the electric field exerts an unbalanced force on the adsorbent particles, which drives them to move along the electric field gradient in the solution. Those particles near the electrodes were first captured and trapped by the electrodes. Due to the strong electric field in the cross-wire areas of the mesh electrodes, coupled with a possible polarization induction effect between adjacent adsorbent particles, other particles close to the wire junctions were polarized and subsequently trapped by the electrodes. In this way, continuous capture of the adsorbent particles would occur, so more and more adsorbent particles would be trapped and deposited on the electrodes (see Fig. 8). This mechanism is supported by the working principle of DEP—a technique that has been used to manipulate polarized particles suspended in fluid media in non-uniform electric field (Khashei et al., 2015). In the case of a spherical particle, the DEP force F_DEP is given by the equation below: (1) $$F_{\text{DEP}}=2\text{π}{R}^3{\text{ε}}_{\text{m}}\text{Re}\left[K\left(\text{ω}\right)\right]\nabla E^2$$ Where the real part of Clausius–Mossotti factor, Re[K(ω)], is defined as: (2) $$K\left(\text{ω}\right)=\frac{{\text{ε}}_p^{*}-{\text{ε}}_m^{*}}{{\text{ε}}_p^{*}+2{\text{ε}}_m^{*}}$$ Where R denotes the radius of the particle, ∇E the magnitude of the electric field gradient, ε_p^* the complex permittivity of the particle, and ε_m^* that of the media. A non-uniform electric field is necessary to induce a DEP force as stated in Eq. (1) (otherwise ∇E = 0). Positive values of Re[K(ω)] denote the induction of a positive DEP force that causes particles to be trapped in the regions of high electric field gradient. Negative values of Re[K(ω)] denote negative DEP, which means the particles would move towards the regions of low or no electric field.

Effects of processing factors on the removal efficiency

As can be seen from Fig. 2 (case 1), the adsorption by zeolite was better than bentonite. The NH₃-N adsorption mechanism by zeolite is mainly chemical adsorption and ion exchange. With the increase of NH₃-N concentration, more NH₃-N is available for exchange, which enhance the capacity of exchange reactions (Zhang et al., 2003).

During the DEP process, it was observed that the removal efficiency was largely enhanced when titanium mesh electrodes were employed in the DEP, resulting in much better efficiency (cases 2 and 3 in Fig. 2). Besides, the stainless steel mesh could be electrolyzed slightly during the DEP process. This clearly indicates that DEP can significantly help improve the removal of NH₃-N as long as a suitable and durable material is used for the electrodes. Based on this initial test result we subsequently chose zeolite as the adsorbent, and titanium mesh as the electrodes, with which we could determine the best possible working conditions for the system.

A phenomenon that the removal efficiency of NH₃-N increased continuously in an adsorption process only with the increase of the zeolite dosage (Fig. 3) may be straightly understood from the increased adsorption sites of zeolite particles in the system (Wan et al., 2017). We also supposed that two factors caused the optimal zeolite dosage during DEP process in Fig. 3. When in the absence of DEP, the removal rate improved with the increase of zeolite dose, which increase adsorption sites. During the DEP-enhanced adsorption, the adsorbed ion number per unit mass of zeolite could reach highest at the dose of 2 g/L which makes the highest efficiency of DEP. And after the dose of 2 g/L, more adsorbents may weaken polarization ability by DEP process. So any further increase of the amount of adsorbent would lead to a reduced removal efficiency.

It can be seen from Fig. 4 that in the lower voltage region (≤15 V), the removal efficiency was increased with the increase of the voltage. As described in Eq. (1), with the increase of applied voltage, the enhance of electric field gradient increased the DEP force. However, after the voltage reached a certain point, the removal efficiency decreased. This can be explained by DEP principle. Different kinds of particles have their own characteristic voltages, in which the positive DEP could happened strongly. Over the characteristic voltage, the particle would leave the original trapped region (Chen et al., 2009; Lapizco-Encinas et al., 2005). In this work the positive DEP would happen when the voltage was 15 V. This speculation is also consistent with the phenomenon observed by Kim et al. (2004) in a diamond suspension.

Because DEP force was short-distance force, the largely enhanced removal efficiency was due to the fact that with the increase of the processing time, more and more zeolite particles could have chances to be carried to the electrodes by flowing water and trapped on the electrodes, and removed from the solution. By calculating the flowing rate and the device size, the wastewater would flow through all the electrodes and be treated in 20 min. So the removal efficiency of NH₃-N almost reached the highest in 20 min (Fig. 5). It is also worthy to note that in comparison with a similar adsorption-only result (Wang, Wang & Zhou, 2012), where it took about 1,440 min to reach the adsorption equilibrium of NH₃-N by zeolite, our adsorption-DEP approach took only 30 min to reach the equilibrium, thus a much faster removal speed. In addition, our approach achieved a significantly higher removal efficiency compared with the adsorption-only method.

It is seen from Fig. 6 that the removal process by adsorption-DEP was particularly powerful at lower initial concentrations: the lower the concentration, the higher the removal efficiency. Figure 6 also shows the difference of the removal efficiency between adsorption-only process and adsorption-DEP process under identical experimental conditions.

Analysis of zeolite particles by SEM

As seen from Figs. 7A and 7B, the broken-down of the zeolite particles resulted in a significant increase of the particle surface area, thus the number of active adsorption sites to NH₃-N. This may explain why the DEP-assisted process had significantly enhanced the removal of NH₃-N in comparison with the adsorption-only method.

Figures 7C and 7D could be understood from the mechanism of DEP rather than electrophoresis. Most of the adsorbent particles were trapped on the cross-wire areas of electrodes where the electric filed gradient was the strongest, indicating that positive DEP took place in this work. Zeolite particles first trapped onto the electrode were polarized again by the electrical field. They produced secondary non-uniform electric field. More particles nearby can be attracted to them because of particle–particle interactions (Ai & Qian, 2010; Cui et al., 2015; Doh & Cho, 2005; Hossan et al., 2013; Jones, 2003), which is illustrated schematically by Fig. 8. Therefore, the NH₃-N removal efficiency could be improved.