Development and uncertainty evaluation of an accurate method for countingliquidborne particles for establishment of particle number concentrationstandards

摘要

National or international standards of particle number concentration in liquids that can beused to calibrate liquidborne particle counters has not yet been established. In the present study, a newhigh accuracy particle counting technique has been developed for the purpose of establishing theprimary standard of particle concentration in liquids. And the uncertainty evaluation of this methodwas executed.In order to count liquidborne particles with high accuracy, a method to distinguish particlesfrom bubbles is required. We employed a flow cytometer for this purpose, and sample particles dopedwith fluorescence dye were used to examine the generation rate of bubbles. While both scattered lightand fluorescence are detected from dye-doped particles, only scattered light is detected from bubbles,and hence we can count particles distinguishably from bubbles. With this method, conditions ofsample preparation for particle counting under which the generation rate of bubbles was reduced weresearched. In case of 10 m diameter particles, a condition under which the concentration of bubblesrelative to that of particles was less than 2.01 % was established.The variation of particle number concentration caused by sampling of solutions wasevaluated. Particle precipitation in solutions causes a concentration distribution. To disperse particleshomogeneously, the container of the sample solution was irradiated with ultrasonic wave. Solutions intest tube each having 2 ml or 1 ml volume was then sampled from the container. While particlecounting was carried out, particles still precipitate and cause the fluctuation of density distribution. Inorder to avoid counting errors due to particle precipitation in test tubes, all the particles contained ineach test tube were counted. The variation in particle number among the test tubes was thus evaluated.In order to reduce bubbles in the tubes, the irradiation was carried out in vacuum.The total uncertainty of the particle number concentration measurement was evaluated bycombining three uncertainty components; the variation due to sampling, the effect of bubbles, and theuncertainty of measuring the volume of a solution in a test tube. The expanded uncertainty of 5.45 %(k = 2) was established.In order to examine the validity of the uncertainty of particle counting by the flowcytometer method, the particle counts obtained by this method was compared with those by themicroscopic method. The entire solution that passed through the flow cytometer was filtered throughan isopore membrane filter, which had 400 nm pore size. Particles were collected onto a circular areaof 4 mm diameter. Total number of particles on the filter was counted to prevent sampling errors,because the number of particles was not large enough to avoid stochastic variations. In case of 10 mparticles, the difference in counts by the flow cytometer method and the microscope method thusfound was about 5 % in average.