the age of the aerosol. The photometric data was reduced by the Mie scattering theory on a computer and showed good agreement with the theory for the evaporation of aerosols developed for this study. The apparatus and a portable version of the same are being adapted to characterize volatile aerosols. The particle size measurement range is 0.5 to 50 μm, and the concentration range is 102 to 106 particles/ cm3. The aerosol is monitored in a controlled atmosphere without disturbance. The aerosol sampling rate is of the order of 1 to 5 liters/ minute. The apparatus is currently under development and final specifications are yet unavailable. However, the cost of the apparatus is pected to be quite low, as most of the parts are simple and easily available. DISCUSSION ILAN CHABAY: Was most of the work you did with monodisperse systems? RANADE: We worked with polydispersed systems which had a standard deviation of about 1.6. CHABAY: What sort of size range do you use? RANADE: Essentially one micrometer. GEORGE SINNOTT: Did you build your own mobility analyzer? RANADE: Yes, Dr. Earl Knutsen built it. He has a differential mobility analyzer with a current measuring system. MILTON KERKER: I have two questions. What is the holdup time and what is the concentration? And the second question, how do you handle the evaporation? Do you have to worry about the Kelvin effect and also I assume you consider the evaporation from a single particle. How do you handle a situation where you have a collection of particles? RANADE: In response to the first question, the holdup time is the order of one to two seconds. That is a a rather short time for coagulation to take place. The concentration is the order of 105 to 106 per cubic centimeter which is not very high for micrometer size particles. Now for the second for the second question, the Kelvin effect has been taken into account in the theoretical equation. One is essentially concerned with the vapor pressure of the water in equilibrium with the droplets. This gives rise to a solution effect and a Kelvin effect and both have been taken into account in solving the equations for the particle size. 39 KERKER: There was an Aerosol Conference in Swan Sea last year, and the question of evaporation figured very centrally. As I recollect some people found that evaporation rates were much lower than they would have calculated from a simple hydrodynamic treatment of a single particle. Some people attributed this to impurities on the surface, but the thing that impressed me more than this was the fact that nobody seemed to come up with even a rough idea as to how the evaporation rate for a single particle would compare to that for a particle in an aerosol mixture. wonder if you have any thoughts on that. I RANADE: Yes, in an aqueous aerosol, evaporation is contributing water vapor to the system so the humidity is changing. For my aerosol at one micrometer size and at a concentration of 106 the contribution to the humidity wasn't very large. I found that it could be treated the same as single particles. Now if you had a denser system, one would have to build something like a cell model. I don't know whether that is justified or not, but a cell model means that instead of assuming that the particle is surrounded by an infinite media, we consider an average interparticle distance and solve the equations accordingly. You can show for this concentration the correction for evaporation will be very low. KERKER: There is one thing that always bothers me with water aerosols. You say the evaporation doesn't contribute very much to the humidity, conversely you need fantastic humidity control in order to define just what is happening with respect to evaporation. RANADE: Yes, depending on which end of the scale you are working. For example, my calculations were based on 106 particles per cc. Assuming complete evaporation of one micrometer particles, the contribution to the relative humidity was one percent. On the other hand when the relative humidity exceeds 85 percent the growth curve is very steep, and a small change can be quite important. REG DAVIES: Dr. Morton Corn at Pittsburgh has been doing some work with sulphuric acid aerosols in in the range of 95 percent relative humidity. Here he observes an enormous change in growth corresponding to a change of 1/2 percent relative humidity, so humidity, so this control of humidity is an important factor. WENDELL ANDERSON: I think the same phenomenon was reported 20 years ago by LaMer. DAVIES: We've done a literature search on this and there have been quite a number of reports about size growth versus humidity, especially for things like sodium chloride aerosols in lungs which have been used to investigate lung humidity. CHABAY: We did some work on cloud droplet growth and we saw some rather strange results at very low levels of supersaturation. The exact values of supersaturation were extremely critical in determining growth rate of the droplets falling through a temperature gradient. NATIONAL BUREAU OF STANDARDS SPECIAL PUBLICATION 412 360° SCATTERING DIAGRAMS FROM INDIVIDUAL AEROSOLS IN A FLOWING STREAM* Thomas R. Marshall, Charles S. Parmenter, and Mark Seaver Department of Chemistry ABSTRACT Nearly complete 360° scattering diagrams have been obtained from each of a large number of individual aerosols which stream in an air sheath through a laser beam. A computer analysis of diagrams from a "monodisperse" aerosol (polystyrene spheres with mean diameter near 1.2 μm) shows that both diameter and refractive index can be determined from each scattering diagram to compile a statistical characterization of the aerosol. Diameter precision is about 0.4 percent and the refractive index can be determined to within 0.7 percent. The method is clearly a practical approach to routine characterization of aerosol sprays containing spherical particles whose diameters range from about 0.3 μm to at least 20 μm. The refractive index determination will reveal differences between particle composition and bulk composition of the parent material as well as composition changes during aging of aerosol sprays. Key words: aerosol light scattering; aerosol size measurements; aerosol spectrometer; aerosol sprays; laser light scattering by aerosols; refractive index; scattering diagrams; 360° scattering by particles. INTRODUCTION Resolved 360° light scattering diagrams from individual particles contain sufficient information to determine the size, the index of refraction, and, in principle, the shape of regular aerosol particulates. Such data offers an attractive route for characterization of aerosols, but unfortunately an experimental barrier exists. Most instruments capable of these measurements require levitation of a particle for periods of of at at least minutes while 360° scans are made [1-5], so that *Contribution No. 2389 from the Chemistry Department, Indiana University. This work has been supported by NSF Grant GP 38275. acquisition of data on statistical numbers of particles in an aerosol is not routinely practical. Professor Frank Gucker and his colleagues at Indiana University have developed an instrument that obviates this problem [6,7]. They have reduced data collection times by four orders of magnitude so that 360° diagrams can be made while particles stream through a laser scattering beam. In this report we describe the initial calibrations and trials of this instrument using spherical particles of about one um diameter. Our first results show clearly that the instrument represents a practical approach for statistical characterization of an aerosol comprised of spherical particles of unknown diameter (0.5 to 20 μm) and unknown refractive index. The Instrument Details of the instrument's optical design and mechanical construction are given elsewhere [7]. A schematic of its operating principle is given in figure 1. In short, an air stream containing aerosol particles intercepts a He-Ne laser beam at one of the focii focii of an elliptical mirror which directs a 360° slice of scattered light to a photomultiplier at the other focus. Angular discrimination is introduced by a 5° rotating aperture between aperture between the scattering plane and the photomultiplier. A nearly complete 360° scattering diagram is obtained in about 20 msec (7° are missing either side of Ŏ° and 180°). The particle is effective stationary in the laser beam during this time. diagrams are displayed on an oscilloscope and can also be stored on magnetic tape by an interface with an XDS Sigma II computer. EXPERIMENTAL SCATTERING DIAGRAMS Scattering "mono All of the data described in this report have been obtained from polystyrene spheres in Dow aerosol sample LS-1028-E. This is a disperse" aerosol with mean particle size reported by the manufacturer to be 1.099 .0059 μm and with a refractive index of 1.59. Figure 2 shows an experimental scattering diagram obtained from a single particle as it streamed through the laser beam. The diagram is plotted from data in computer storage. The mirror symmetry of the scattering pattern is especially significant because it proves to be a sitive indication of an accurate scattering diagram. A test of the reproducibility of the data is offered in figure 3 where four scattering curves from the same particle are compared. These curves were obtained in successive scans as the particle traversed the 1.5 mm cross section of the laser beam. Such reproducibility is typical of the instrument. In a size analysis by computer (see below), these four curves were assigned to particles with diameters 1.180, 1.180, Figure 1. A schematic of the 360° scattering instrument. An aerosol stream intercepts a He-Ne laser beam ("vertical" polarization) at one of two foci of an ellipsoidal mirror. Light scattered in the "horizontal" plane is intercepted by a segment of that mirror and directed to a photomultiplier at the second focus. A rotating aperature allows only a 5" segment of the scattered light to be detected at any given time by the photomultiplier. The aperature rotates at about 3000 rpm so that a 360° scan is completed in 20 msec. The photomultiplier signal comprising a comprising a 360° scattering diagram if fed through a logarithmic amplifier to a computer for storage and processing. |