A fast and accurate method of using electrical mobility scans for the direct measurement of aerosol charge distributions
M. Wild, J. Meyer, G. Kasper
Journal of Aerosol Science, 2012, 52 (Oct. 2012), 69-79
We present an efficient and accurate method of determining the aerosol concentration within a well-defined range of electrical mobilities – such as distinct peaks in a mobility spectrum – directly from a continuous DMA scan. The method applies most immediately to aerosol charge distribution measurements via a tandem DMA arrangement. It is based on maintaining a constant transfer probability by moving the transfer function across the spectrum at such a rate that the time window during which particles of an arbitrary mobility can pass through the DMA remains constant. (The volume DMA flow rates remain constant as well.) This implies a specific ramp function for the scan voltage, which we derive, thereby establishing a direct and simple link between the cumulative number of particles exiting the DMA while scanning across a peak, and the concentration enclosed within that mobility range at the inlet. We derive the method and show also that it is a special case of a more general theory (Knutson & Whitby, 1975a) for calculating arbitrary moments of a mobility spectrum. By relaxing the goal of obtaining a complete size distribution via continuous mobility scan (Wang & Flagan, 1990) in favor of the lesser requirement of measuring only the total concentration within certain, well-defined bounds of mobility, one gains both computational efficiency and accuracy (because fewer measurements are required), as well as some sensitivity in detecting very small peaks (demonstrated down to less than 0.01% of the total particle concentration).
The method is validated against a more conventional approach based on fitting appropriate mathematical functions (log-normal distributions) to the individual charge peaks of the same mobility spectra. Tandem DMA measurements were performed using laboratory generated aerosols of 60 and 135 nm in Boltzmann charging state. It is shown that the conventional method introduces some error due to the requirement of an independent determination of the neutral fraction. If this is corrected for, measured charge fractions agreed to within 1–3%.