Retrospective Analysis of Two Decades of Particle Wall-loss Correction from the UCR Dual-90 m3 Collapsible Chamber – A New Dynamic Size-independent Coagulation-corrected Method

Chen Le, YANYU ZHANG, David R. Cocker III, University of California, Riverside

     Abstract Number: 405
     Working Group: Aerosol Physics

Abstract
Minimizing uncertainty associated with particle wall-loss corrections are critical for accurately quantifying aerosol yield from environmental chamber experiments. Typically, the experiment-specific, size-independent particle wall-loss rate was derived by fitting the SMPS-measured total particle number decay during the final three hours of the experiment. Particle wall-loss in the UCR previous-generation dual 90-m³ collapsible chambers were enhanced by the charged surfaces of chamber walls. It was previously observed that particles in the chamber tend to approach a charge steady-state during the course of an experiment, resulting in dynamic changes of particle charge distribution as well as the electrostatic enhancement on particle-wall deposition. These resulted in size-insensitivities as well as day-to-day variation of measured electrostatic-enhanced particle wall-loss rates in UCR collapsible chambers. Particle number loss due to particle-particle coagulation could also bias the derivation of particle wall-loss rates, especially when increased particle number loadings were present in the chamber. These observations suggest that real-time particle wall-loss rates do not remain constant during a chamber experiment and therefore use of single wall-loss rate may not accurately describe particle wall-loss over the duration of an experiment. A coagulation-corrected, time-resolved particle wall-loss correction method was warranted to improve wall-loss correction estimates.

A particle coagulation dynamics model was developed in this work to update the particle wall-loss correction for experiments using the UCR collapsible chambers. The model was designed to calculate dynamic particle wall-loss rates after accounting for coagulation and these experiment-specific dynamic values were used to accurately account for both coagulation and particle dynamics. The model was embedded into a MATLAB program to efficiently calculate the particle wall-loss corrections of over 1900 experimental datasets from UCR collapsible chamber experiments conducted during the past two decades. The results verified dynamic changes of particle wall-loss rates within individual experiments, possibly attributed to the changes of particle size, particle charge distribution, chamber surface charge along with chamber geometry. The retrospective summarizes the difference of particle wall-loss rates calculated using traditional and updated correction methods across different years and generations of chamber bags. The particle wall-loss rate discrepancies between the different correction methods grew with increasing particle number loadings. Particle number loss due to coagulation was observed to have a significant impact on the total particle number loss during the first few hours of the experiment when higher particle number loadings were present. Overall, minimum of ~ 5% and ~ 15% overestimation of the final corrected particle volume are determined when coagulation is not accounted for in experiments with at least 10⁴ cm^-3 particles and 5 × 10⁴ cm^-3 particles, respectively. Preliminary wall-loss corrected data for a series of repeated experiments shows that even maintaining constant initial injections goals, small subtleties still cause measurable bias regarding the generated SOA.