868 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS INTRODUCTION During the past few years increased attention has been given to the particle size distribution of aerosol products. While the importance of this property has been recognized for many years in relationship to oil- based insecticides, it has not been deemed very important for other aerosols except those intended for oral use. Although particle size alone does not determine efficacy, it does give an indication of it in such products as disinfectants, underarm deodorants and antiperspirants, feminine hygiene sprays, hair sprays, and other aerosols. A product de- veloped for application to a given area would be troublesome if the particles were so small and light that more particles became airborne than were deposited onto the desired surface. On the other hand, a room deodorant would not be very effective or efficient if the particles settled rapidly following each application. Not only can improper par- ticle size distribution be detrimental to the product, it can present a potential toxicity problem if the airborne particles are inhaled. The danger of inhaling mineral oil particles and the possibility of lipoid pneumonia occurring, especially in infants, is well known. Dautrebande (1-3) and Brown (4) report that particles of 1-5 • and smaller are easily inhaled and reach the respiratory tract. Many methods for the determination of particle size distribution are available however, direct microscopic measurement, light scattering, microphotography, and impaction technics have found the greatest application. Several of these methods have previously been reviewed (5). The Cascade Impactor* has been found useful in classifying par- ticles produced from an aerosol spray in a range of from about 0.5 to 32 • in diameter. Sciarra et al. (5) studied the use of the Cascade Impactor for determining the particle size distribution of several hair spray for- mulations and different aerosol valves. One of the problems noted in us- ing the Cascade Impactor is the determination of the amount of material deposited upon each ot• the slides. In the above study, a weight-by- difference method was used and, while found to be acceptable, was time consuming and limited to those aerosols which contained a nonvolatile solid. It is the purpose of this investigation to study the use of the Cascade Impactor for the determination of the particle size distribution of sev- Scientific Advances, Inc., Columbus, Ohio 43212.

PARTICLE SIZE DETERMINATION 869 eral aerosol products having a different particle size distribution. Par- ticular attention has been devoted to methods which can be used to determine quantitatively the amount of material deposited on each slide. The results obtained from the weight-by-difference method and an- other method will be compared. EXPERIMENTAL A Cascade Impactor, Model CI-S-6, was used throughout the study. This model consists of a set of six round glass slides approximately 38 mm in diameter, fitted at each of the small jet openings. The instrument has been calibrated and the different sized particles will impact upon the glass slide as follows: Upper slide 32 v Next lower 16 tx Next lower 8 v Next lower 4 tx Next lower 2 t• Bottom slide 1 t• Filter O. 5 • The impactor was assembled in the verticle position with the cham- ber having the largest diameter opening on top and each additional lower chamber decreasing in opening size. A glass expansion and sampling chamber was then fitted over the upper opening as shown in Fig. 1. Each of the glass slides was previously weighed using a semimicro analytical balance. The aerosol preparation was accurately weighed and a sample of about 10 g was sprayed intermittently into the chamber. Prior to introduction of the sample, a vacuum of 17 in. of mercury was applied so that a flow rate of 12.5 1./min was obtained. This setting was used throughout the study. The aerosol container was then reweighed in order to obtain the sample weight. Vacuum was applied for an addi- tional 5 min, after which time the vacuum was discontinued and the Cascade Impactor was disassembled. Each glass slide was then allowed to dry and, using the semimicro balance, was reweighed to obtain the weight of residue remaining at each stage. Knowing the amount of nonvolatile material present in the aerosol originally, one can calculate the amount ot• product deposited on each slide. This procedure was used for those samples analyzed by the weight-by-difference method.