The effect of width ratio of collection nozzle to acceleration nozzle on the classification performance has been studied by using a rectangular jet virtual-impactor. Numerical calculations showed that smaller cut-size would be obtained if the impactor was set with larger nozzle width ratio. The sharpness of the classification was also expected to be higher for larger nozzle width ratio. The experimental results showed, however, that if the nozzle width ratio was larger than about 1.7, cut-size data disagreed with the theoretical values. The data for the sharpness showed similar tendency.
 Flow visualization experiments have been carried out in order to find out the causes of the disagreement. It was found that the flow in the impactor of large nozzle width ratio was separated from the wall of the collection nozzle and large eddies were generated there. Some of the particles separated into the collection nozzle would be caught by the eddies and exhausted to the separation zone, resulting in larger cut-size and less sharp separation than expected by the numerical calculations.
 The ratio of undeflected flow rate to total flow rate was increased, and clean air flows were introduced so as to eliminate the eddies. Then smaller cut-size and sharper classification were obtained.

View full abstract

Direct detection of airborne particles is very useful for various studies such as clean rooms, air leakage of filters, pinhole testing, etc. This technique covers the weak points of the commonly used optical measuring equipments, which have field disturbance with sampling, particulate adhesion inside the sampling tube, etc. In this paper, the methods and experiments are reported for remote detection of particles. Scattered laser light of a particle is remotely detected by using an optical system, a video system and an image processing system. By emitting a He-Ne laser beam into the clean room space, scattered light of a particle is caught by a TV camera through optical lenses located at about one meter distance from the detected space. The TV image of scattered light is recorded on a video tape. The recorded image is analyzed to automatically obtain the number of particles in any objective space. Furthermore, an example of application of this technique is discussed for the adhesion to the surface of silicon wafers.

View full abstract

A series of experiments was carried out to prepare the ultrafine particle of titanium oxide using gas phase chemical reactions. The effect of reaction conditions, such as the kind of the oxidizing agent, oxygen or water vapor, the partial preessure of TiCl₄ vapor and the reaction temperature, was investigated.
 The identification, mean size, size distribution and shape of particles were studied by X-ray diffraction and T. E. M observation.
 The products were identified as titanium dioxide both anatase and rutile, white in color, several tens to hundreds of nm in size, spherelike and flakelike in shape. The average size of oxide particles becomes smaller with smaller P_TiCl4 and higher temperature.

View full abstract

Atmospheric concentrations of rare earth elements and their size distributions were measured at various sites in the Osaka area by applying the Andersen sampler and instrumental neutron activation analysis. Except La, rare earth elements measured were mostly in coarse particles above 1.5 μm in diameter. La was found to exist considerably in fine particles below 1.5 μm in diameter, and its recent size distribution shows rather the bimodal pattern at many sites. The ratio of La to Sm is almost always greater than 20 in fine particles. This ratio had been about 10 at least until 1987, and it seemed to change between 1979 and 1983.
 Various source samples such as soil and road dust, particles from coal-fired plants, oil-fired plants and oil refineries were analyzed elementally by instrumental neutron activation analysis, and concentration ratios of La to other rare earth elements and to V were derived from obtained analytical results. Concentration ratios of rare earth elements in soil and road dust and in particles from coal-fired plants were quite similar to their abundance ratios in crust. As for particles from oil-fired plants, concentrations of rare earth elements were generally lower than those in crustal particles, and the concentration pattern was also different from the crustal abundance pattern. The concentration ratio of La to V in particles from oil-fired plants was below 0.003. This is below one tenth at least when compared to the reported values in the United States. Particles from oil refineres such as the catalyst reclaiming tower and CO-boilers contained concentrated rare earth elements originating from the zeolite catalyst, and it is suggested that only about 100 ng/m³ emissions from these plans may distort the concentration pattern of rare earth elemets to the similar degree observed in the Osaka area.
 Almost all rare earth elements in coarse particles seem to originate from crustal particles, because the concentration pattern of rare earth elements was quite similar to these from soil and read dust samples. On the other hand, contributions of crustal particles and emissions from oil-fired plans were small for fine particles, and the concentration pattern of rare earth elements suggested a probability of oil refinery emissios. Recently, Ce and La are consumed in large quantities in catalyst for automobiles. Judging from the present consequences, the distortion of concentration pattern of rare earth elements in fine particles is inferred to be mainly due to emissions from catalyst automobiles rather than those from oil refineris.

View full abstract