Environ. Sci. Technol. 2007, 41, 857-862
Hydrogen Cyanide Exhaust Emissions from In-Use Motor Vehicles MARC M. BAUM,* JOHN A. MOSS, STEPHEN H. PASTEL, AND GREGORY A. POSKREBYSHEV Department of Chemistry, Oak Crest Institute of Science, 2275 East Foothill Boulevard, Pasadena, California 91107
Motor vehicle exhaust emissions are known to contain hydrogen cyanide (HCN), but emission rate data are scarce and, in the case of idling vehicles, date back over 20 years. For the first time, vehicular HCN exhaust emissions from a modern, in-use fleet at idle have been measured. The 14 tested light duty motor vehicles were operating at idle as these conditions are associated with the highest risk exposure scenarios (i.e., enclosed spaces). Vehicular HCN was detected in 89% of the sampled exhaust streams and did not correlate with instantaneous air-fuel-ratio or with any single, coemitted pollutant. However, a moderate correlation between HCN emissions and the product of carbon monoxide and nitric oxide emissions was observed under cold-start conditions. Fleet average, cold-start, undiluted HCN emissions were 105 ( 97 ppbV (maximum: 278 ppbV), whereas corresponding emissions from vehicles operating under stabilized conditions were 79 ( 71 ppbV (maximum: 245 ppbV); mean idle fleet HCN emission rates were 39 ( 35 and 21 ( 18 µg-min-1 for cold-start and stabilized vehicles, respectively. The significance of these results is discussed in terms of HCN emissions inventories in the South Coast Air Basin of California and of health risks due to exposure to vehicular HCN.
Introduction Hydrogen cyanide (HCN) is a ubiquitous trace gas, but its global atmospheric budgets remain poorly understood (1), despite nearly 25 years of research (2). Both HCN and acetonitrile (CH3CN) are thought to play minor roles in the nitrogen cycle (3), and HCN can interfere with measurements of total reactive nitrogen (NOy) (4). Typical mean HCN mixing ratios in the unpolluted lower troposphere (at 0-2 km) are around 200 pptV (parts per trillion, pptV, 1 pptV ) 1 part in 1012 by volume or moles) (1). Ocean uptake appears to be the primary HCN sink, with corresponding tropospheric lifetimes of 5.3 months according to recent estimates using a global three-dimensional model analysis (1). Other, minor sinks of atmospheric HCN are thought to include oxidation by OH and O(1D) as well as photolysis in the stratosphere, leading to related atmospheric residence times of about 2.5 years (2). Cicerone and Zellner argue that wet scavenging is a minor sink (2). It is widely accepted (1, 5, 6) that biomass burning constitutes the dominant global source of HCN (0.1-3.2 TgN-yr-1) and CH3CN (0.2-1.1 Tg-N-yr-1) in the free tropo* Corresponding author phone: (626)817-0883; fax: (626)817-0884; e-mail:
[email protected]. 10.1021/es061402v CCC: $37.00 Published on Web 12/13/2006
2007 American Chemical Society
sphere, and, therefore, both are used routinely as atmospheric tracers of biomass burning (7). Hydrogen cyanide emissions from motor vehicles generally are considered to make a negligible contribution to global sources (1). However, this assumption is based on limited experimental data. Voorhoeve et al. reported that nitric oxide (NO) could be reduced in the presence of carbon monoxide (CO) and molecular hydrogen (H2) over platinum automotive exhaust catalysts in the 400-800 °C temperature range to produce HCN according to eq 1 (8).
NO + CO + 5/2H2fHCN + 2H2O
(1)
Sua´rez and Lo¨ffler observed the surface-mediated formation of HCN from ammonia (NH3) and methane (CH4) on platinum (9), rhodium, and iridium (10).
NH3 + CH4 f HCN + 3H2
(2)
These results raise the concern that HCN may be generated in vehicle emissions control systems. Previous dynamometer studies (11-15) examined the dependence of HCN emission rates on vehicle characteristics, emission control technology, and driving conditions. Harvey et al. used these early data, and others, to estimate fleetaveraged HCN emission rates of 12.1 mg-km-1 (16), for a fleet consisting exclusively of light duty motor vehicles (LDMVs) equipped with a three-way catalyst (TWC) emission control system. Reports on vehicular HCN emissions from a modern fleet are limited. Becker et al. employed Fourier transform infrared (FTIR) absorption spectroscopy to measure HCN emission rates from a fleet of 22 vehicles (cars and trucks), which were tested on a dynamometer (17). However, no HCN emissions were observed above their instrumental detection limit of 2 mg-km-1. Karlsson reported on HCN emissions from a fleet of five LDMVs tested under two driving cycles on a chassis dynamometer (18). One vehicle, a 1989 Volvo equipped with a first generation catalyst, exhibited high HCN emissions, 8.2 or 11.7 mg-km-1 depending on the test cycle; this vehicle also was the highest NH3 emitter (14.0 or 36.5 mg-km-1) from the fleet of five. Since there are no current reports on HCN emissions by idling vehicles, we carried out a study to sample a modern, in-use fleet of LDMVs and to determine HCN emission rates under normal idle conditions. Measurements were made under cold-start (“cold”) and under stabilized (“warm”) conditions. The real-world significance of these data are discussed and were used to derive emission inventories within the regional context of the South Coast Air Basin (SoCAB) of California. In addition, the health risks associated with vehicular HCN exhaust emissions were re-examined for the first time Since Harvey et al.’s study (16) 20 years ago.
Experimental Methods Test Vehicles and Fuels. Tailpipe emission measurements on a fleet of 14 vehicles (Table 1) were carried out at Oak Crest’s parking lot between August 24 and September 1, 2004. All vehicles ran on regular, unleaded, Phase 2 California Reformulated (summer) gasoline (fuel sulfur content
