Environ. Sci. Technol. 2001, 35, 4660-4667
Direct Measurement of OH Radicals from Ozonolysis of Selected Alkenes: A EUPHORE Simulation Chamber Study MANFRED SIESE,† KARL H. BECKER,‡ KLAUS J. BROCKMANN,‡ H A R A L D G E I G E R , * ,‡ ANDREAS HOFZUMAHAUS,† FRANK HOLLAND,† DJURO MIHELCIC,† AND KLAUS WIRTZ§ Institut fu ¨ r Chemie und Dynamik der Geospha¨re, Institut II: Tropospha¨re, Forschungszentrum Ju ¨ lich, D-52425 Ju ¨ lich, Germany, Bergische Universita¨t Gesamthochschule Wuppertal, Fachbereich 9, Physikalische Chemie, Gaussstrasse 20, D-42097 Wuppertal, Germany, and Centro de Estudios Ambientales del Mediterraneo (CEAM), Parque Tecnologico, Calle Charles Darvin, 46980 Paterna Valencia, Spain
Reactions of ozone with alkenes can be a significant source of hydroxyl radicals in the atmosphere. In the present paper, the formation of OH radicals in the ozonolysis of selected alkenes under atmospheric conditions was directly observed. The experiments were carried out in the European photoreactor EUPHORE (Valencia, Spain). OH radicals were quantitatively detected by means of laser-induced fluorescence (LIF) using a new analytical instrument, which has been constructed on the basis of an existing setup already established in field studies. The OH radicals observed resulted directly from the reaction of ozone with the corresponding alkene. There was no indication that OH radicals were produced in the system by secondary processes. The experimentally observed concentration-time profiles of OH and ozone were excellently described by chemical modeling using explicit reaction mechanisms. The following OH yields were derived: 2,3-dimethyl-2-butene: (1.00 ( 0.25); 2-methyl2-butene: (0.89 ( 0.22); trans-2-butene: (0.75 ( 0.19); R-pinene: (0.91 ( 0.23). In addition, the experiments carried out were modeled using the Regional Atmospheric Chemistry Mechanism (RACM), an established condensed chemical model applied in tropospheric chemistry. For 2,3dimethyl-2-butene, 2-methyl-2-butene, and trans-2-butene the calculated concentration-time profiles of OH and ozone are in quite good agreement with the experimental data. However, in the case of R-pinene, the model fails for the simulation of OH due to the high grade of mechanism condensation, which results in a poor characterization of the primary reaction products. * Corresponding author phone: +49-202-439-3832; fax: +49-202439-2757; e-mail:
[email protected]. † Institut fu ¨ r Chemie und Dynamik der Geospha¨re, Institut II. ‡ Bergische Universita ¨ t Gesamthochschule Wuppertal. § Centro de Estudios Ambientales del Mediterraneo (CEAM). 4660
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 23, 2001
Introduction The main source of OH radicals in the atmosphere is the photolysis of ozone in the presence of water vapor. However, the formation of OH from reaction of ozone with alkenes is of considerable interest, as it may significantly influence the radical budget in urban and rural environments (1). Under certain conditions, the reaction may be the predominant OH radical source at night. Though the mechanism of the gas-phase ozonolysis of alkenes has been intensively studied, considerable uncertainties still exist concerning the OH radical yield as well as the detailed mechanism of the reaction sequence. The mechanism of the gas-phase ozonolysis is commonly believed to proceed in analogy to the Criegee mechanism, which was primarily proposed for the liquid phase. First step of the reaction chain is the addition of ozone to the double bond of the alkene (R1R2CdCR3R4), forming a primary ozonide (mole ozonide). The subsequent cleavage of this primary ozonide leads to a carbonyl compound (R1R2CO) and an excited intermediate, often referred as excited Criegee biradical (R3R4OO*). It is assumed that this excited biradical is either stabilized by collision or undergoes further decomposition, producing free radicals such as H and OH as well as stable molecules. In most studies the yield of OH radicals has been estimated from the concentrations of stable products observed, from the decay of OH scavengers added to the ozone/alkene mixture or from products of the reaction of OH with the scavenger (2-5). Scha¨fer et al. (6) used two different OH scavengers added to the ozone/alkene mixtures. However, the relative decay rates of these scavenger molecules were inconsistent with the OH kinetics expected. In conclusion, the yield of OH radicals in the ozonolysis of alkenes still remains uncertain. Very recently, Fenske et al. (7) investigated the formation of OH radicals from the ozonolysis of selected alkenes as a function of total pressure in the range of 200760 Torr for the first time. These studies, performed using the scavenger technique, exhibited a significant increase of the OH yield with decreasing total pressure for ethane and propene. On the other hand, no pressure dependence of the OH yield was observed for the ozonolysis of higher alkenes (C g 4). Only a few direct measurements of OH radical formation in the ozonolysis of alkenes are available (8-10). Donahue et al. (8) investigated the reaction of ozone with ethene at low pressure, including direct measurement of the OH concentration. The authors report an OH yield of (40 ( 20)%, which is much larger than the value of (20 ( 2)% determined from indirect measurements by Mihelcic et al. (10). These authors converted the OH radicals formed in the ozonolysis to HO2 via reaction with CO. The HO2 concentration was then measured using matrix isolation/spin electron resonance (MIESR). Recently, Pfeiffer et al. (9) observed the formation of OH radicals by UV absorption spectroscopy in the ozonolysis of several terpenes under atmospheric conditions. Recently, Kroll et al. (11) investigated the ozonolysis of selected alkenes using direct detection of OH by laser-induced fluorescence and obtained pressure-dependent OH yields for trans-2-butene, 2,3-dimethyl-2-butene, trans-3-hexene, 3,4-dimethyl-3-hexene, and trans-4-octane. However, most of their experiments were carried out at total pressures below 100 Torr (down to 1 Torr). In another very recent paper (12), these authors performed additional RRKM/master equation calculations with respect to their experimental investigations (11). 10.1021/es010150p CCC: $20.00
2001 American Chemical Society Published on Web 11/02/2001
a
b
FIGURE 1. a: The European photoreactor EUPHORE in Valencia, Spain. The mounting position of the LIF instrument is marked by the white circle. b: Schematic setup (laser beam plane) of the LIF instrument used for the OH radical measurements in the EUPHORE chamber (CM1,2 ) remote-controlled cardanic-mounted mirrors, PSD A,B ) position-sensitive UV photodiodes). In the present work, OH radicals and OH yields from the ozonolysis of selected alkenes were determined directly using the laser-induced fluorescence. The experiments were carried out in an outdoor simulation chamber at atmospheric pressure in the absence of sunlight.
Experimental and Modeling Techniques The EUPHORE Chamber. The ozonolysis studies were carried out in the EUPHORE outdoor simulation chamber (13-16). This facility (see Figure 1a) was mainly designed for the investigation of smog systems. A protective housing (not shown in Figure 1a) offers also excellent possibilities for studies to be performed in the dark. Briefly, the chamber consists of a half-spherical FEP bag (fluorine-ethene-propene, foil thickness 127 µm) with a volume of about 187 m3 and a base diameter of 9.2 m. Purified and dried ambient air is used to fill the chamber and flushing it between the experiments. The air purification system consists of a compressor, a condensate trap, and an absorption dryer filled with a suitable molecular sieve, leading to a dew point below 213 K. In addition, a special charcoal adsorber unit eliminates NOx and reduces oil vapor and nonmethane hydrocarbons below 0.3 µg m-3. However, CO and CH4 are not removed, and their concentrations remain at ambient levels. Rapid mixing of the reactants (
