A Cost-Effective Electrostatic Precipitator for Aerosol Nanoparticle Segregation

Copyright 2015 American Association for Aerosol Research


INTRODUCTION
Measuring the size of aerosol particles having diameters smaller than 100 nm (i.e., nanoparticles) is important for assessing their environmental impacts (McMurry 2000) and investigating their potential technological applications (Biskos et al. 2008). The most effective way for sizing aerosol nanoparticles is by classifying them based on their electrical mobility using differential mobility analyzers (DMAs; Knutson and Whitby 1975). Despite that DMAs can classify particles with a high resolution by simply changing the potential difference between their two electrodes, their high cost and bulky size is limiting for widespread applications. Diffusion batteries (DBs) that distinguish aerosol particles based on their diffusivity, can also be used as particle size classifiers (DeMarcus and Thomas 1952). Although compact designs of DBs have been proposed, a number of technical limitations (McMurry 2000) has made them less favorable compared to electrical mobility classifiers.
In this letter, we introduce a simple and cost-effective method for size segregating aerosol nanoparticles by employing tubes composed of Electrostatic Dissipative Materials (EDMs). EDM tubes have surface resistivities that range from 10 5 to 10 12 V/sq. Applying a potential difference along the tube creates an electric field of varying strength that has a radial and an axial component within the tube (see the discussion section). This field affects the charged particles passing through the tube in two ways: the axial field decelerates the particles and therefore increases their residence time in the tube and their chance for diffusional deposition to its walls, whereas the radial field removes particles by electrostatic deposition. As a result of these two processes, EDM tubes can be viewed as a combination of a DB and an electrostatic precipitator (or a crude DMA), with the advantage of being significantly more simple and inexpensive.
The relative penetration efficiency of the particles passing through an EDM tube (i.e., the ratio of the particle number concentration at the outlet when a potential difference is applied along the tube over that when it is grounded) can be predicted using a modified version of the semi-empirical model employed for diffusion batteries: Here a, b, g, and d are positive empirical constants, whereas m diff and m elec are dimensionless parameters accounting for particle deposition by diffusion and electrostatic forces, respectively. For laminar flow conditions, m diff is given by (Hinds 1999): Here, D is the diffusion coefficient of the particles, L is the length of the tube, and Q is the aerosol volumetric flow rate. In a similar manner, m elec can be defined as: where Z p is the electrical mobility of the particles and f g is a factor accounting for the dependence of the edge effects of the electric field to geometrical parameters, given by: Here d tube is the inner diameter of the EDM tube and L HV is the distance between the grounded inlet and the position along the tube where the high voltage is applied. Expressions for the parameters used to calculate D and Z p are given in Section S1 in the online supplemental information (SI).

EXPERIMENTAL
The EDM tube (Freelin Wade, Model 1A-405-81) used in our tests had a length of 240 mm and inner diameter of 6.4 mm. Three metallic rings were attached along its length as shown in Figure 1: one at the inlet, one at the outlet, and one at an intermediate point between the inlet and the outlet. The metal rings at the inlet and the outlet were grounded, while the intermediate ring was connected to a positive high-voltage power supply (Fug, Model HCN14-12 500) that can deliver up to 12.5 kV.
The experimental setup for characterizing the classifier consisted of an aerosol Spark Discharger Generator (SDG; Schwyn et al. 1988), a custom-made DMA (Section S2 in the SI) with a closed recirculating system for the sheath flow, the EDM tube and two Condensation Particle Counters (CPCs; Agarwal and Sem 1980) as shown in Figure 1. In brief, the SDG was used to produce polydisperse singly-charged silver particles of variable sizes by adjusting the energy per spark and the carrier gas flow as described by Tabrizi et al. (2009). Positively-charged, monodisperse particles having electrical mobility diameters from 10 to 55 nm were obtained after classification by the DMA. The monodisperse particles were then passed through the EDM tube before their concentration was measured by an ultrafine CPC (uCPC; TSI Model 3025).
The relative penetration efficiency through the EDM tube when the applied voltage at the intermediate electrode varied from 1 to 8 kV was calculated by: Here N(0) and N(V) are the average particle number concentrations measured over 60 s by the uCPC downstream the EDM tube when zero and V volts were respectively applied at the intermediate electrode. A second CPC (TSI Model 3072) was used to verify that the particle number concentration upstream the EDM tube was stable throughout the measurements. Figure 2 shows the measured and predicted relative penetration efficiency of particles having electrical mobility diameters from 10 to 55 nm that enter the EDM tube. Measurements are shown for two different aerosol flow rates, 0.32 and 0.76 lpm, when 1 to 8 kV were applied on the intermediate electrode of the tube. The semi-empirical model (i.e., Equation (1)) was fitted to the measurements using the least squares method, yielding the following constants: a D 0.65, b D 2668.83, g D 0.35, and d D 4.21. For all particle sizes and operating conditions tested, the agreement between predictions and measurements was within 15%, which is in the same range to the uncertainty of our setup.

RESULTS AND DISCUSSION
For a fixed aerosol flow, the larger particles require higher potential differences between the intermediate ring electrode and the two ends of the EDM tube in order to decrease their penetration probability. When the potential difference is also fixed, the relative penetration of the particles increases logarithmically with their size, yielding curves that become steeper as particle size decreases (Figure 2). The size resolving capability of the EDM tube is higher in these steep regions of the curves; a feature that is highly desirable when electrostatic precipitators are used for particle segregation. Another feature of the EDM tubes that makes them attractive classifiers is that in the steep regions of the penetration efficiency curves, the relative size resolution is almost constant. This is verified by the fact that the curves are almost parallel when a logarithmic horizontal axis is used in the plots shown in Figure 2.
Interestingly, the relative penetration efficiency curves of the EDM tube are steeper compared to those of classical parallel-plate precipitators having the same cross-sectional area (Section S3 and Figure S1 in the SI), indicating that they would perform much better as classifiers. This characteristic can be attributed by the diffusional deposition of the charged particles, which is enhanced by the axial component of the FIG. 1. Schematic diagram of the experimental setup and the EDM-tube classifier (inset). Metal rings are attached to the tube for grounding the inlet and the outlet, and for applying a high voltage at an intermediate point (i.e., 40 mm downstream the inlet) along the tube. A DMA was used to select monodisperse positively charged silver particles produced by a SDG. The monodisperse particles having electrical mobility diameters ranging from 10 to 55 nm were then passed through the EDM tube, which was operated at voltages ranging from zero to 8 kV. The particle number concentration was measured downstream the EDM tube by a uCPC. A CPC was used to check the stability of the number concentration of the monodisperse particles upstream the EDM tube. THE EDM-TUBE ELECTROSTATIC PRECIPITATOR v electric field that can significantly decelerate the smaller charged particles in the EDM tube. To further investigate the role of the electric field on the penetration of the particles through the EMD tube, we used a numerical model to calculate its strength when 1 kV is applied at the intermediate ring electrode (Section S4 and Figure S2 in the SI). The electric field is significantly distorted near the three electrode rings, inducing a radial component that is extremely strong in their vicinity (i.e., the "Hot Spot" regions indicated in Figure S2), but decays fast along the axial and radial dimension. This component of the field attracts the positively charged particles towards the walls right at the edge of the grounded electrode ring at the inlet of the tube (i.e., "Hot Spot 1") thereby increasing the probability of their deposition. At the same time, the mean strength of the axial component of the electric field upstream (Zone 1) and downstream (Zone 2) the high voltage electrode is 30 and 5.5 kV/m, respectively ( Figure S2c). For 10-nm singly charged particles, this field can reduce the mean convective particle velocity by 0.07 m/s in Zone 1 and increase it by 0.01 m/s in Zone 2. For typical flow rates used through the EDM tubes (i.e., in the range of an lpm), this can substantially increase the residence time of the particles, if not force them to move opposite to the direction of the flow, in Zone 1 thereby increasing their probability of deposition to the tube walls. This feature is reflected by the relative penetration efficiency curves that are noticeably steeper compared to those of parallel plate precipitators as discussed above.

CONCLUSIONS
In this work, we demonstrate the ability of EDM tubes to be used as effective and compact electrostatic precipitators that can be employed for size segregation of charged aerosol nanoparticles, and provide a semi-empirical model to predict their performance. These tubes can be considered a combination of a DB and an electrostatic precipitator. Compared to DBs, where only the flow rate can be adjusted to select particles of different size, EDM tubes offer higher flexibility as they can segregate particles by simply adjusting the potential difference along their inner surface. Compared to classical parallel plate electrostatic precipitators, EDM tubes exhibit steeper cut-off curves. This feature is highly attractive for their potential use in mobility spectrometers or in tandem systems downstream a DMA for measuring size-dependent particle properties such as hygroscopicity and volatility. Considering also their simple design, high portability, and negligible cost, EDM tubes can open up new opportunities in environmental monitoring of aerosol nanoparticles as discussed above.

SUPPLEMENTAL MATERIAL
Supplemental data for this article can be accessed on the publisher's website.