EXPERIMENTAL VALIDATION OF THE MULTI WAVELENGTH LIGHT EXTINCTION TECHNIQUE ON POTASSIUM CHLORIDE NANOPARTICLES

Nowadays, the control of the particle size distribution and concentration of nanoparticles or nanoparticle aggregates gained special industrial and scientific interest. In this contest, my research project regarded the experimental validation of the Multi Wavelength Extinction Technique (MWLE) on an aerosol of Potassium Chloride nanoparticles. The MWLE is an optical technique that allows in situ real-time monitoring nonintrusively and is able to carry out simultaneous concentration and size measurement in the nanometric size range. In the frame of the present work, a facility for the aerosol generation and for allowing accurate light extinction measurements has been build. As the main goal of this research project was to validate the MWLE, the developed experimental setup allowed also the simultaneous measurement of the same particle flow with both the Muti Wavelenght Light Extinction and with a Scanning Mobility Particle Sizer (SMPS) instrument. The MWLE requires a specific, regularized inversion algorithm for the retrieval of the particle distribution in volume V-PSD, so both the experimental and numerical aspects of the technique were investigated. The Tikhonov NNLS method, with smoothing matrix 0th and 2nd order discrete derivative operator, was applied and the L-curve method (used for finding the optimal regularization parameter) was optimized. All the inversion results gave a volume concentration around one order of magnitude bigger than the one given by the SMPS. Finding the parameter influencing this discrepancy is still a subject of investigation. For this reason normalized distributions, with normalization made with respect to the peak of the distribution, are considered. The comparison with the SMPS, in terms of Normalized Particle Volume Distributions, resulted in good agreement for both the methods, but the 0th order appeared more sensitive to the inversion parameters, like the diameter intervals in which the inversion was applied. Thus the 2nd order was chosen for investigating the effect on the particle size and concentration of the operating parameters like atomization pressure and the ejector feed pressure, which leads to different pressures in the test chamber. Good agreement was found between the MWLE and the SMPS results, in terms of Normalized Particle Volume Distributions and of Volume Mean Diameter D4,3. The measurements showed that, decreasing the atomization pressure, the particle size increases and their concentration decreases. Furthermore, the MWLE showed high sensitivity to the measured transmittance, so an accurate measurement procedure was requested. Summarizing, the objectives of the project have been both the development of the above-mentioned experimental setup and the validation of the MWLE algorithm using a commercial sizing system.