Dynamic Characteristics of a Fast-Response Aerosol Size Spectrometer

Abstract

The fast integrated mobility spectrometer (FIMS) is a highly sensitive instrument with a fast response time. The time response of the FIMS is limited by the mixing of the aerosol in the inlet of the instrument, which “smears” the detected aerosol over a range of time. The response time is also limited by the different transit times that particles experience in the instrument due to differences in particle classification trajectories. Experiments show that the difference in particle transit times can be corrected using a simple model of the particle trajectories. Furthermore, the “smearing” effects in the inlet can be corrected by de-convolving the time series of particle counts (the temporal de-convolution) in each size bin before inverting the data. The dynamic response of the FIMS was investigated by measuring an aerosol subjected to a step-change and sinusoidal-change in number concentration. The attenuation of the FIMS signal, without using the temporal de-convolution, was measured with and without an aerosol neutralizer in the FIMS inlet. Due to its large internal volume, the neutralizer significantly slows down the time response of the FIMS when installed in the inlet. For a sinusoidal signal at 0.33 Hz, the measured attenuation of the FIMS without the neutralizer was 56% versus an attenuation of 82% when the neutralizer was used. When the temporal de-convolution was applied at the same frequency, the FIMS was able to capture the full variation of the aerosol size distribution; however the random noise in the derived size distribution was amplified.

This work was supported by the Office of Biological and Environmental Research, Department of Energy (DOE), under Contract DE-AC02-98CH10866, and the Laboratory Directed Research and Development program at the Brookhaven National Laboratory (BNL). BNL is operated for the DOE by Battelle Memorial Institute. Jason Olfert also acknowledges partial support from the Goldhaber Distinguished Fellowship from Brookhaven Science Associates. The authors also wish to acknowledge Dr. Peter Takacs for his help with the optics on the FIMS.

Notes

1. Ideally, the flow rate in the neutralizer would be 5 L/min to ensure that the aerosol had sufficient residence time to become fully neutralized, yet short enough to minimize mixing. However, due to the limited selection of orifices on hand, the nearest flow rate that could be achieved in the neutralizer was 5.32 L/min. This higher than recommended flow rate may result in the aerosol not achieving complete charge equilibrium, which might result in a small shift in the measured distribution. However, this is not important in this work since the absolute value of the measured distribution is not important, but rather the measured change in the distribution.

2. Quant et al. (1992) report a “time to 5%” of 0.8 s for step-change down in the input concentration. This equates to a time constant of 0.27 s for a first-order instrument. The step-change was produced by sampling ambient and particle-free filtered air through an electrically-operated three-way valve.

3. Buzorius (2001) measured the frequency response of the CPCs by using two DMAs. The first DMA was fixed to provide a steady-state triangular aerosol distribution, which was then classified by a second DMA, with a narrower transfer function, whose voltage was sinusoidally oscillated between the peak concentration and a lower concentration.