V777x7777a

Electronics Repair Manuals

Schematic Diagrams and Service Manuals

Get Instant Access

After-

Stage N

After-

To Vacuum Pump

FIGURE 11.57 Principle of operation of cascade impactor (adapted from Marple and Willeke, 1979).

To Vacuum Pump

FIGURE 11.57 Principle of operation of cascade impactor (adapted from Marple and Willeke, 1979).

Thus, the impaction efficiency should be greatest for larger, denser particles and higher flow velocities. The factors involved in particle impaction on surfaces are discussed in detail by John (1995).

There are two overall types of impactors in widespread use: cascade and virtual impactors.

Cascade impactors. Impactors have been used to obtain different size fractions of ambient particles in the range of diameters ~ 0.5-30 /um. The range can be extended down to 0.05 pm by operating some of the later stages at reduced pressures (Hering et al., 1978, 1979). The cascade impactor, as its name implies, is a series of impactor plates connected in series or in parallel (Fig. 11.57). The diameters of the nozzles or slits above each impactor plate become increasingly smaller as the air moves through the impactor so that the air moves increasingly faster through these orifices and smaller and smaller particles impact on the plates [see Eq. (R)].

Impactors with various designs as well as different types of impaction surfaces are in use (e.g., see Chow, 1995). Examples include the Lundgren impactor, the Anderson sampler, the Mercer impactor, and the Uni versity of Washington Mark III impactor. An impactor that is in wide use is the MOUDI (Microorifice t/ni-form Deposit /mpactor) (Marple et al., 1991). This device collects particles down to 0.056 pm in aerodynamic diameter and, as the name implies, gives a uniform particle deposit on the plates. This uniform deposit helps in carrying out chemical analysis by such techniques as X-ray fluorescence. The uniformity in deposition is obtained by using multiple nozzles located at specific distances from the center of the impactor plates and rotating the plates beneath the nozzles.

Two problems with particle collection by impactors are bounce-off and reentrainment (John, 1995). Reen-trainment is the resuspension of a previously collected particle from the surface into the gas flow due either to the motion of the air over the surface or to impact of an incoming particle. When a particle strikes a surface, if it does not stick, it can bounce off back into the gas steam, break into fragments, or cause a previously adsorbed particle to be knocked off into the gas stream; in all three cases the collection efficiency is lowered and the net effect is referred to as bounce-off. To minimize these problems, the surface of the impactor is often coated with a soft, energy-absorbing substance such as oil, water, grease, resin, or paraffin, which helps to absorb the kinetic energy of the striking particle; a summary of the types of agents used to minimize bounce-off and reentrainment is given by Marple and Willeke (1979), Cahill (1979), and Turner and Hering (1987).

While the use of soft surfaces would seem to be mandated by the foregoing discussion of bounce-off problems, there are a number of disadvantages to coating the impactor surfaces with a substance such as grease. For example, it makes accurate mass determinations difficult and can introduce such a large background of certain chemicals that the chemical analysis of these elements in the particles becomes difficult. In addition, with such surfaces one cannot use chemical analytical techniques that only probe the upper surface layer because the coating surrounds some of the collected particles.

Virtual impactors. The virtual impactor is a modified type of impactor, an example of which is shown in Fig. 11.58; one commonly used type of virtual impactor is known as the dichotomous sampler. The basis of virtual impactors is that the airstream impacts against a mass of relatively still air rather than against a plate. The inertia of the particles carries them into the still air

Sampler inlet

Sampling tube Annular slit

Sampling tube Annular slit

Sampler inlet

Filter' Aluminum

To vacuum pump

FIGURE 11.58 Schematic diagram of a virtual impactor (adapted from Conner, 1966).

Filter' Aluminum

To vacuum pump

FIGURE 11.58 Schematic diagram of a virtual impactor (adapted from Conner, 1966).

mass, which is slowly withdrawn through a filter to collect the particles. This type of impactor avoids the problem of particle reentrainment from the impaction surface caused by air motion over the collected particles or by dislodging due to collisions of incoming particles with the impactor surface. It also avoids the problem of bounce-off or of using greases that may interfere with subsequent chemical analysis.

c. Electrostatic Precipitators

Electrostatic precipitators operate on the principle of the attraction of a charged particle for an oppositely charged collector. They have been used for both collecting particles for further analysis and for controlling particulate emissions from sources. In one common design, the particles in air can be charged if introduced into a cylindrical chamber containing a wire down the axis of the cylinder that is at a high negative voltage (e.g., 5-50 kV) relative to the walls of the chamber. A corona discharge is set up around the wire and this produces ions; the negatively charged ions are attracted to the positively charged outer walls. These ions collide with the particles in the air, charging them and causing them to move to the outer walls to be captured there. In place of the corona discharge, ions may also be generated using radioactive bombardment of the particles.

While electrostatic precipitators have relatively high collection efficiencies (99-100%) over a wide range of particle sizes 0.05-5 ¡xm), there are a number of disadvantages. These include the lack of size information, particle reentrainment due to sparking, and practical problems such as high cost and shock hazards. As a result, they have not been widely used in ambient air studies.

An example of a study in which this approach was applied involved the use of a transmission electron microscopy (TEM) grid as the collector plate in the electrostatic precipitator (Witkowski et al., 1988). After sample collection, analysis by TEM (vide infra) could then be carried out.

A related area is that of single-particle levitation, which has been used in a number of studies to isolate a single particle and study its properties (e.g., see papers by Tang and co-workers in Chapter 9). A review of this area is given by Davis (1997).

d. Sedimentation Collectors

These collectors are used primarily for large particles (> 2.5 pm), that is, those in the coarse particle range. They include collection by gravitational sedimentation (e.g., dustfall jars) as well as by centrifugal sedimentation, which allows collection in the sub-micrometer range (e.g., centrifuges and cyclone collectors).

Gravitational sedimentation only collects the large particles that settle out of the atmosphere fairly quickly. This dustfall generally consists of particles that are relatively large and, as such, are not particularly relevant to the focus of this book. Thus dustfall collectors will not be discussed further.

The principle of centrifugal collection is, of course, well known. Collection of particles using centrifugation involves passing the aerosol at a controlled rate through a rapidly spinning air mass. Collection of particles in ranges as small as ~ 0.1-1 ¿im has been reported using this technique. The cyclone collector, a modification of the centrifuge technique, is based on bringing the air samples into a stationary cylindrical vessel at high velocity; a vortex is formed by the entry of the air tangential to the length of the vessel and particles in this vortex are subjected to a centrifugal force that depends on their size (Fig. f 1.59). As a result, particles of different sizes are deposited at different locations along the length of the cyclone separator. Although cyclone collectors have been applied to size distribution measurements by using a series of cyclones in parallel, each having a different cut size, they are most commonly used as precollectors to remove larger particles (~3- to 30-/Lim diameter) before the air sample enters a device such as an impactor designed for the measurement of particles in smaller size ranges.

Air out

Air out

2. Measurement of Physical Characteristics: Mass and Size a. Mass

The total mass of particles per unit volume of air is one of the major parameters used to characterize particles in air and, along with size, is the basis of air quality standards for particulate matter (see Chapter 2). Methods of mass measurement include gravimetric methods, /3-ray attenuation, piezoelectric devices, and the oscillating microbalance.

(1) Gravimetric methods The most straightforward method of determining the particle loading of the atmosphere is to weigh a collection substrate such as a filter before and after sampling. However, care must be taken to be sure that both temperature and relative humidity are carefully controlled when weighing both the loaded and clean substrate. As discussed earlier, some filters and/or the collected particles are hygroscopic and unless care is taken to equilibrate them at a fixed temperature and relative humidity, the change in water content may completely mask the change in mass due to the particles. In addition, problems such as forces due to static electricity on the filter that interfere with accurate weight measurements must be controlled. Finally, particulate loading can change the sampling air flow rate and lead to large errors in determining the actual volume of air sampled.

(2) fi-Ray attenuation /3-Particle beams (electrons) emitted from a radioactive source are attenuated when they pass through a filter on which particulate matter has been collected. (¿8-particle beams rather than a-particle beams or y-rays are used because a particles do not penetrate typical thicknesses of filter well and y-rays are too penetrating and hence would require large sample thicknesses.) Figure 11.60 shows a

|3-ray detector

Filter with particulate matter p-ray source

FIGURE 11.60 Schematic diagram of a typical /3-ray attenuation device for measuring particulate mass.

schematic of a /3-ray attenuation device, which consists essentially of a ¡3 source such as 14C, a ¡3 detector, and a means of positioning the filter paper containing the particulate matter between the two. The ratio of the transmission of /J-rays through a clean and loaded portion of the filter, respectively, is related to the particle loading via a Beer-Lambert type of relationship:

/0 and I are the intensities of the /3-rays that have passed through the clean and loaded portions of the filter, respectively, X is the thickness of the deposit, and ¡x is an attenuation constant that is approximately proportional to the density (p) of the material deposited. The mass per unit area deposited on the filter, given by pX, is the parameter desired in this measurement. Rearranging Eq. (S), one obtains ln(/„//) = (ix/p)pX. (T)

The parameter p/p is a constant known as the mass absorption coefficient; with the assumption that this is independent of the type of absorbing particles (an assumption that generally holds well enough to cause < 10% uncertainty), the value of In(/„//) is directly related to the parameter of interest, pX = mass per unit area.

Such measurements can be carried out on filters with different cutoff sizes to obtain size resolution as well (e.g., see Spagnolo and Paoletti, 1994).

(3) Piezoelectric microbalance The piezoelectric microbalance is a resonant frequency device. The piezoelectric effect is the development of a charge on some crystals such as quartz when a stress is applied; the stress may be mechanical (e.g., added weight) or elec trical. Such crystals may be used as part of a resonance circuit to provide very stable, narrow-band frequencies;

the quartz crystal is plated on two sides with a thin conducting layer and leads are connected to the resonance circuit so the crystal replaces an LC network. The obtained frequency of vibration (t>0) depends on a number of parameters of the crystal but is usually

~5-10 MHz. However, if a mass (Am) becomes attached to one side of the crystal, it changes the resonant frequency by an amount At',, such that

as long as the increase in mass Am is much smaller than the mass (m) of the active part of the crystal.

Particulate matter from ambient air can be deposited on the crystal in various ways, for example, by using it as an impaction device. The mass of the collected particles can then be determined by following the change in the frequency. Alternatively, a reference crystal held at the same temperature and pressure as the crystal on which the particles are collected can be used, and the difference in frequencies between the two crystals can be determined.

The piezoelectric microbalance is very sensitive, capable of detecting ~10~8-10~9 g. The particles collected on the crystal surface can be chemically analyzed after collection using surface-sensitive techniques. One limitation is possible overloading of the crystal; thus when the collected mass reaches ~ 0.5-1% of the mass per unit of the crystal, the surface must be cleaned.

(4) Oscillating microbalance The tapered-element oscillating microbalance is based on a similar principle to the piezoelectric microbalance. A hollow glass piece is mounted with the wider end fixed and a filter attached to the narrower end. The tip oscillates at a particular frequency in an applied magnetic field. As particles are collected on the filter, the resonant frequency of the glass tends to decrease. A feedback is used to maintain the oscillation frequency and provides a measurement of the collected mass (e.g., see Patash-nick and Rupprecht, 1991). Good agreement has generally been observed between measurements made using the tapered-element oscillating microbalance and Hi-Vol filter methods (e.g., Eldering and Glasgow, 1998).

b. Size

There are several different approaches that are commonly used to determine particle size distributions in air. One of them, impaction, has been discussed earlier. Multistage impactors with different cut points are used extensively to obtain both mass and chemical composition data as a function of size for particles with diameters > 0.2 p,m. Others, including methods based on optical properties, electrical or aerodynamic mobility, and diffusion speeds, are described briefly in the following section. The condensation particle counter (CPC) is used as a detector in combination with some of these size-sorting methods.

The reader is cautioned to keep in mind that atmospheric particles are not all spherical nor even necessarily simple in shape. Thus, as discussed in Chapter 9.A, the term size cannot be uniquely defined for atmospheric particles. As a result, a measurement of the distribution of sizes using an impactor that is based on inertial characteristics, for example, may not give the same results as a size measurement based on optical techniques that use light scattering. With this caveat in mind, let us examine the most commonly used methods of determining the size distribution of atmospheric particles.

(1) Optical methods Optical counters, optical microscopy, and electron microscopy fall under this heading. A review of optical methods is given by Baron et al. (1993).

Single-particle optical counters. These instruments are used to measure particles in the ~0.1- to 10-/j,m range by measuring the amount of light scattered by a single particle (Martens and Keller, 1968). As discussed in Chapter 9.A.4, the amount of this Mie scattering depends not only on the refractive index but also on the radius of the particle; hence the intensity of scattered light is a measure of the particle size. Assuming that the particles are spherical, smooth, and of known refractive index, one can calculate, using Mie theory, the intensity of scattered light of wavelength A at various angles (6) to the incident beam for a particle of a given size. Integrating over all scattering angles and wavelengths (since "white" incandescent sources are normally used in these instruments), one obtains the theoretical response of the single-particle counter, that is, the curve of scattered light intensity as a function of the particle diameter. Typical theoretical response curves are shown in Fig. 11.61 (Cooke and Kerker, 1975).

Calibration of these single-particle counters is usually carried out using monodisperse polystyrene latex or polyvinyl latex spheres, which are available in sizes from ~0.1 to 3 p,m and have a refractive index of f.6; alternatively, aerosols with lower refractive indices may be generated from liquids such as dioctyl phthalate (m = 1.49). Whitby and Willeke (1979) discuss the

Was this article helpful?

0 0

Post a comment