AAAR 35th Annual Conference October 17 - October 21, 2016 Oregon Convention Center Portland, Oregon, USA
Abstract View
Performance of Different Particle Wall-loss Correction Methods for Aging Experiments of Alpha-pinene SOA in a Smog Chamber
Ningxin Wang, Neil Donahue, SPYROS PANDIS, Carnegie Mellon University
Abstract Number: 694 Working Group: Instrumentation and Methods
Abstract The interaction of particles with the chamber walls often represents a major complication in the analysis of the results of smog chamber experiments. The uncertainty introduced by the corresponding data corrections increases with the duration of the experiment and is especially significant in chemical aging experiments. A number of particle wall-loss correction methods have been used in experiments focusing on secondary organic aerosol (SOA) yields. These include uses of size-independent first-order loss rate constants, size-dependent corrections, corrections based on the ratio of organics to sulfate in the chamber, etc. While all of these approaches can work well for some types of experiments, they may not be suitable for chemical aging experiments that last several hours and take place under different chamber conditions (e.g., lights are turned on or off).
In this work, we evaluate the performance of several particle wall-loss correction methods with regard to aging experiments of alphapoi-pinene ozonolysis SOA conducted in the Carnegie Mellon University smog chamber (10 m3 Teflon chamber). In these experiments the SOA is produced during the dark reaction of alpha-pinene and ozone, and then the products are exposed to OH produced in the chamber by the photolysis of HONO. We show that the size independent and the organic/sulfate ratio approaches often introduce significant errors in aging experiments. The size dependent correction performs better but the corresponding rate constants may change from experiment to experiment and even during an experiment when the conditions change. An approach is proposed for the characterization of these losses in the different phases of the experiment and the corresponding wall loss corrections.