When a battery is not actually a battery
10/01/2002
By Robert P. Donovan
Past columns referred to the use of "diffusion batteries" to measure the concentration of aerosol particles in cleanrooms. Because these devices are not well known in today's cleanroom circles, let's consider how they could become more important in classifying the properties of future cleanrooms designed for manufacturing smaller and more contaminant-sensitive precision products.
The term "diffusion battery" is deemed by many outside of the aerosol technology community—and even some within it—to be a misnomer and unnecessarily confusing. A diffusion battery has no electrodes, generates no voltages and has none of the features that an electrical engineer or even the common man in the street associates with the term battery.
Rather, it is a device that separates aerosol particles according to size, using the size dependence of particle diffusion coefficients as the mechanism affecting the separation.1 Used in combination with a condensation particle counter, it is useful for collecting particle-size data in the sub-0.2-micron particle size range and into the size ranges well below the capability of optical particle counters.
The action of a diffusion battery can be understood by considering aerosol flow through a tube. In aerosol flow through a long cylindrical tube, diffusion causes aerosol particles to depart the streamlines, traverse the boundary layers, contact the tube walls and be captured. Small particles have larger diffusion coefficients than large particles and are more rapidly removed from the aerosol flow than large particles.
Thus, the size distribution of the aerosol particle population in the stream shifts toward larger particles as the aerosol stream progresses down the length of the tube.
Knowing the volume flow rate of the aerosol stream and the length of the tube lets you determine the diffusion coefficient of a monodisperse aerosol from measurements of the entering and exiting particle concentrations. Typically, a downstream condensation particle counter (CPC) is used to measure particle concentration, because the diffusion battery itself does not count particles—it just separates them according to size.
 Two models of diffusion batteries with one tap of the foreground battery connected to a condensation particle counter. |
null
The aerosol stream enters one end of the diffusion battery, flows axially along its length and exits at the opposite end. Taps along the flow path allow particle sampling at multiple points, each tap representing a different tube length at which particle concentration data can be collected.
Such multiple taps allow the size distribution of polydisperse aerosols to be approximated. These taps can be sequentially connected to a CPC. Using just a single CPC to serially collect particle penetration data eliminates the detector variation that would be introduced by using multiple CPCs, but the aerosol properties must remain constant throughout the time period required to take samples from all taps.
Alternatively, operating a diffusion battery of one fixed length (no taps) at different volume flow rates can also approximate the size distribution of a polydisperse aerosol. The accuracy of this approach also depends upon the assumption that the aerosol size distribution remains the same at each flow velocity sampled.
An interesting and valuable property of the diffusion battery is that particle loss by diffusion does not depend on tube diameter as long as the volume flow rate through the tube remains constant. As tube diameter increases, the slower linear flow velocity compensates the greater distance that particles must traverse perpendicular to the flow axis to contact the tube walls and be captured.
Similarly, a single long tube can be cut into numerous small lengths, say 100, and the aerosol stream can be fed to these 100 short lengths arranged in parallel. The linear flow velocity through each short section is now just 1/100 of that through the single long tube. In turn, the net particle capture by diffusion of the parallel array is the same as that of the single long tube in which the velocity of the aerosol stream is 100 times faster.
This feature has important design advantages in that the performance of a high-resolution, but impractically long, tube can be duplicated by a honeycomb flat disk composed of short lengths of that tubing in parallel. Some of the most practical diffusion battery structures use wire screens to simulate this geometry.
The advantage of the diffusion battery/condensation particle counter combination is its capability to separate particles by diffusion coefficient—and thus by size—in the particle size range below the sensitivity limits of optical particle counters.
Next month's column will present data collected in a cleanroom using such a combination and comparing particle size distribution and concentration in a cleanroom at rest with that of the same cleanroom in a typical operational mode.
Robert P. Donovan is a process engineer assigned to the Sandia National Laboratories and a monthly columnist for CleanRooms magazine. He can be reached at [email protected].
Reference
1. Knutson, E. O., "History of Diffusion Batteries in Aerosol Measurements," Aerosol Science and Technology, 31 (2,3), Aug/Sept 1999, p.83.