Environ. Sci. Technol. 2005, 39, 4823-4832
Products of Ozone-Initiated Chemistry in a Simulated Aircraft Environment
removal rate constant was 6.3 h-1 when T-shirts were not present, compared to 11.4 h-1 when T-shirts were present.
Introduction ARMIN WISTHALER,† GYO ¨ N G Y I T A M AÄ S , ‡ DAVID P. WYON,‡ PETER STRØM-TEJSEN,‡ DAVID SPACE,§ JONATHAN BEAUCHAMP,† ARMIN HANSEL,† T I L M A N N D . M A¨ R K , † A N D C H A R L E S J . W E S C H L E R * ,‡,# International Centre for Indoor Environment and Energy, Technical University of Denmark (DTU), DK-2800 Kgs. Lyngby, Denmark, Institute of Ion Physics, University of Innsbruck, Innsbruck, Austria, Boeing Commercial Airplane Group, P. O. Box 3707, MC 02-WH, Seattle, Washington 98124, Environmental and Occupational Health Sciences Institute, University of Medicine and Dentistry of New Jersey (UMDNJ) and Rutgers University, 170 Frelinghuysen Rd., Piscataway, New Jersey 08854
We used proton-transfer-reaction mass spectrometry (PTR-MS) to examine the products formed when ozone reacted with the materials in a simulated aircraft cabin, including a loaded high-efficiency particulate air (HEPA) filter in the return air system. Four conditions were examined: cabin (baseline), cabin plus ozone, cabin plus soiled T-shirts (surrogates for human occupants), and cabin plus soiled T-shirts plus ozone. The addition of ozone to the cabin without T-shirts, at concentrations typically encountered during commercial air travel, increased the mixing ratio (v:v concentration) of detected pollutants from 35 ppb to 80 ppb. Most of this increase was due to the production of saturated and unsaturated aldehydes and tentatively identified low-molecular-weight carboxylic acids. The addition of soiled T-shirts, with no ozone present, increased the mixing ratio of pollutants in the cabin air only slightly, whereas the combination of soiled T-shirts and ozone increased the mixing ratio of detected pollutants to 110 ppb, with more than 20 ppb originating from squalene oxidation products (acetone, 4-oxopentanal, and 6-methyl5-hepten-2-one). For the two conditions with ozone present, the more-abundant oxidation products included acetone/propanal (8-20 ppb), formaldehyde (8-10 ppb), nonanal (∼6 ppb), 4-oxopentanal (3-7 ppb), acetic acid (∼7 ppb), formic acid (∼3 ppb), and 6-methyl-5-hepten2-one (0.5-2.5 ppb), as well as compounds tentatively identified as acrolein (0.6-1 ppb) and crotonaldehyde (0.60.8 ppb). The odor thresholds of certain products were exceeded. With an outdoor air exchange of 3 h-1 and a recirculation rate of 20 h-1, the measured ozone surface * Corresponding author. Telephone: 732 235-4114; fax: 732 5301453; E-mail:
[email protected]. † University of Innsbruck. ‡ Technical University of Denmark. § Boeing. # Environmental and Occupational Heath Sciences Institute. 10.1021/es047992j CCC: $30.25 Published on Web 06/03/2005
2005 American Chemical Society
At the typical cruising altitudes of commercial aircraft, the mixing ratio of ozone in the air outside the cabin is significantly higher than that found in the most polluted cities in the world. Ozone levels vary with altitude, latitude, and time of year. During late winter and early spring, at altitudes of >10 000 m and latitudes above 45°, ozone levels in the range of 500-800 ppb are common (1). Even at lower altitudes and in different seasons, ozone mixing ratios are often >200 ppb. It is this air, with its associated ozone, that is used to ventilate the aircraft cabin. Only about half of the wide body commercial aircraft and 0.5 ppb: m/e ) 31 (formaldehyde); m/e ) 59 (acetone/propanal), m/e ) 75 (tentatively identified as propionic acid), m/e ) 101 (4-oxopentanal), m/e ) 109, 127 (6-methyl-5-heptene-
TABLE 3. Relative Abundance of Species Observed during Conditions When Ozone Was Present in the Simulated Aircraft Cabin Relative Abundance [%] chemical or chemical class saturated aldehydes unsaturated aldehydes carboxylic acids squalene oxidation products methanol organic nitrates/peroxyacylnitrates
cabin + O3 cabin + T-shirts + O3 55-60 5 14 19-24 ∼1 ∼1
40-43 4-5 13 37-41 ∼1 ∼1
2-one/octenal), and m/e ) 157 (decanal). These seven ion signals account for 80% of the observed increase relative to the cabin/O3 condition and can be largely explained by the O3-squalene reaction (see below). Acetone and propanal are isobaric and cannot be distinguish by PTR-MS; the same is true for 6-methyl-5-heptene-2-one and octenal. For the cabin + T-shirts + O3 condition, we assumed a propanal mixing ratio of 2 ( 1 ppb and an octenal mixing ratio of 0.35 ( 0.15 ppb, based on expectations from the respective homologous series. Table 3 compares the relative abundance of species observed during the two conditions when ozone was present in the simulated aircraft cabin (i.e., cabin + O3 and cabin + T-shirts + O3). More-Detailed Comparisons among the Four Conditions. In the following paragraphs, we will make comparisons among different classes of compounds that contribute to the total signals shown in Figure 3: saturated aldehydes, unsaturated aldehydes, carboxylic acids, squalene oxidation products, and organic nitrates. For the cabin/baseline or the cabin + T-shirts + no-ozone conditions, the values in the figures to follow should be taken as upper limits, because other unidentified chemicals may contribute to the signal at the m/e’s characteristic of the compounds. However, for the conditions in which ozone is present, the mixing ratio increases are ascribed, with greater certainty, to the identified products of the resulting oxidative chemistry. 1. Saturated Aldehydes. As shown in Figure 4, saturated aldehydes were major contributors to the total PTR-MS signals measured within the simulated aircraft cabin. Propanal is not included in this figure, because we cannot distinguish its signal from that of acetone; this signal is reported below as the sum of acetone and propanal. For both conditions without ozone, the formaldehyde level was ∼3 ppb, while levels of higher-molecular-weight aldehydes were
