Secondary organic aerosol formation from ozone reactions with single terpenoids and terpenoid mixtures

Ozone reacts with indoor-emitted terpenoids to form secondary organic aerosol (SOA). Most SOA research has focused on ozone reactions with single terpenoids or with consumer products, and this paper reports the results from an investigation of SOA formation from ozone reactions with both single terpenoids and mixtures of d-limonene, α-pinene, and α-terpineol. Transient experiments were conducted at low (25 ppb) and high (100 ppb) initial concentrations of ozone. The three terpenoids were tested singly and in combinations in a manner that controlled for their different reaction rates with ozone. The SOA formation was assessed by examining the evolution in time of the resulting number size-distributions and estimates of the mass concentrations. The results suggest that at higher ozone and terpenoid concentrations, SOA number formation follows a linear trend as a function of the initial rate of reaction. This finding was valid for both single terpenoids and mixtures. Generally speaking, higher ozone and terpenoid concentrations also led to larger geometric mean diameters and smaller geometric standard deviations of fitted lognormal distributions of the formed SOA. By assuming a density, mass concentrations were also assessed and did not follow as consistent of a trend. At low ozone concentration conditions, reactions with only d-limonene yielded the largest number concentrations of any experiment, even more than experiments with mixtures containing d-limonene and much higher overall terpenoid concentrations. This finding was not seen for high ozone concentrations. These experiments demonstrate quantifiable trends for SOA forming reactions of ozone and mixtures, and this work provides a framework for expanding these results to more complex mixtures and consumer products.

► Concentrations of SOA formed due to ozone and terpenoid mixtures were examined. ► SOA number formation mostly followed a linear trend with initial reaction rate. ► Higher reaction rates led to larger geometric mean diameters of distributions. ► Higher reaction rates led to lower geometric standard deviations of distributions. ► Mass formation did not follow as consistent of a trend as number formation results.