Further Investigation of Ozone, Carpets and VOCs
Morrison and Nazaroff (2000) conducted laboratory studies to measure ozone uptake on four samples of fresh and aired out whole carpet and on the corresponding carpet fibers and backings. Results were described in terms of reaction probability defined as the rate of ozone loss on the surface normalized by the rate of ozone-surface interactions. For whole carpet and carpet backing, the apparent reaction probability was of the magnitude of 10^-5 to 10^-4 (i.e., one molecule of ozone reacted for every 100,000 to 10,000 molecules hitting the surface) when referenced to the floor area that would be covered by carpet. Reaction probabilities of the order of 10^-7 to 10^-6 were measured on carpet fibers, referenced to the total estimated fiber area. All of the materials tested exhibited aging such that the rate of ozone uptake diminished with increasing cumulative exposure. Estimates indicated that carpet might dominate painted walls and ceilings in scavenging ozone from indoor air, even though the superficial surface area of carpet is smaller than the area of the painted surfaces.

A physically based mathematical model was subsequently developed to describe the mass transport and uptake of ozone on carpeting (Morrison and Nazaroff, 2002b). An existing model of particle deposition from turbulent flow to indoor surfaces was extended to include surface resistance to the uptake of reactive gases. Surface resistance was parameterized in terms of the measured pollutant-surface reaction probability. The model predicted that the deposition velocity of ozone onto carpet should lie in the range of 0.016 to 0.064 cm/s, values that were consistent with field measurements. Owing to its higher reaction probability, carpet backing was predicted to consume approximately the same amount of ozone as the fibers even though the surface area of the fibers was much higher.

Emission rates of aldehydes and other VOCs were measured in the experiments with the ozone exposed and unexposed carpets described above (Morrison and Nazaroff, 2002a). Surface interactions of ozone with the carpets produced formaldehyde, acetaldehyde and various C3 through C13 normal aldehydes as well as several unsaturated aldehydes. Nonanal was the predominant aldehyde. All materials, whether fresh or aired, exhibited strong secondary aldehyde emission factors. Total aldehyde emission factors increased markedly with ozone exposure from 1 to 70 µg/m²-h for unexposed materials to 60 to 800 µg/m²-h for exposed materials. One exposed carpet emitted about 200 µg/m²-h of 2-nonenal with ozone exposure. This compound is of particular significance as it has a very low odor threshold (0.15 ppb for trans-2-nonenal). It was estimated that the emissions of 2-nonenal could persist over a period of years at rates sufficient to impact room odors.