Abstract. Measurements of atmospheric organic nitrates derived from isoprene, i.e., "isoprene nitrates", were conducted from July 14 to August 19, 1998, as part of the 1998 summer intensive measurement campaign of the Program for Research on Oxidants: PHotochemistry, Emissions, and Transport (PROPHET) at the University of Michigan Biological Station in Pellston, Michigan. The measurements were conducted using on-line chromatography in conjunction with a nitrate-selective detection scheme. Measured concentrations of isoprene nitrates ranged from 0.5 parts per trillion (ppt), the detection limit of the method employed, to 35 ppt. In this paper we discuss the contribution of the isoprene nitrates to NOy, which was typically 0.5 -1.5% of total odd nitrogen, but up to -4% for well-aged air. Concentrations of isoprene nitrates exhibited a strong diurnal variation consistent with their expected chemical and physical removal rates. In this work we also discuss the chemistry of the precursor peroxy radicals and the NOx dependence of isoprene nitrate formation.

A recently developed prototype mobile laboratory mass spectrometer, incorporating an atmospheric pressure ionization (API) interface, is described. This system takes advantage of the small size, lower voltage requirements, and tandem MS abilities of the cylindrical ion trap mass analyzer. The prototype API MS uses small, low-power pumps to fit into a 0.1-m 3 self-contained package weighing <45 kg. This instrument has been adapted to allow rapid interfacing to electrospray ionization, desorption electrospray ionization, and direct analysis in real-time sources. Initial data indicate that these techniques provide rapid detection and identification of compounds for quality control, homeland security, and forensic applications. In addition, this instrument is self-contained and compact, making it ideally extensible to mobile laboratory and field analyses. Initial MS and MS/MS data for analyses of drugs, food, and explosives are presented herein. O ver the past 15 years, efforts toward implementing mass spectrometry (MS) in the field have steadily grown with particular focus on environmental, forensic, defense, and security applications [I, 2]. Many of the previously developed fieldable MS instruments use gas chromatography (GC-MS) because of the high degree of confidence obtained using GC retention times and electron ionization (EI) mass spectral matching [3][4][5][6]. The specificity of GC-MS comes at the expense of time, with individual analyses often taking in excess of 10 min (not including sample preparation). Reducing the amount of time required for chemical identification requires separation-free MS. In the absence of chromatography, tandem MS (MS/MS) improves confidence for the identification of individual components in mixtures [7]. With spectral acquisition rates > 1 Hz, atmospheric pressure ionization (API) techniques such as electrospray ionization (ESI), desorption electro spray (DESI), and direct analysis in real time (DART), a fieldable API MS system capable of MS/MS is of high value to environmental, forensic, defense, and force protection agencies. There have been few attempts to construct fieldable MS instruments with an API interface to date [8][9][10][11] because of the difficulty of displacing the gas load of the atmospheric inlet.The cylindrical ion trap (CIT)-a simplified geometry of the hyperbolic quadrupole ion trap capable of Address reprint requests to Dr. Mitch Wells, Griffin Analytical Technologies, LLC, R&D, 3000 Kent Avenue, West Lafayette, IN 47906, USA. E-mail: mitch.wells@icxt.com performing MS n analyses-has been shown to be amenable to miniaturization [12][13][14] while maintaining benchtop-quality performance characteristics. The miniaturized CIT enables a smaller vacuum system with smaller pumps that consume less power, providing a fieldable instrument. Although GC-MS instruments currently dominate the portable mass spectrometry industry, estimates indicate that the confidence for molecule identification by MS/MS is about that provided with GC-MS (when combining MS information w...

Measurements indicate that HONO may play a larger role in the reactive nitrogen budget than previously expected for a rural site. HONO/NO2 for a 24 hour period was observed to be 0.09-0.25 and suggests the likelihood of a significant heterogeneous production pathway or pathways.

[1] We report measurements of isoprene nitrates as part of the Southern Oxidants Study during the summer of 1999 at a rural/forest site in Tennessee. Average midday concentrations of the isoprene nitrates were $115 ppt. This is $10 times greater than the previously reported concentrations during the Program for Research on Oxidants: Photochemistry, Emissions, and Transport (PROPHET) study in 1998 at Pellston, Michigan, representing as much as 5% of NO y . Here we investigate the possible factors for the large difference in concentrations. To investigate the role of the NO x concentration on the isoprene nitrate production chemistry at the two sites, [OH] was calculated using a simple steady state model. The results of this calculation help explain the difference in magnitude of the isoprene nitrate concentrations between the two field sites in terms of the [NO x ]-dependent behavior of the OH-initiated oxidation of isoprene and the subsequent isoprene peroxy radical reactions with NO x . However, it is also clear that the large apparent differences in the photochemical ages of the air masses sampled at the two sites significantly impacted the observed concentrations.

[1] A large set of isoprene and isoprene oxidation product concentration data from four North American sites was examined to assess the NO x dependence of the daytime oxidation of isoprene. Sites that represent a wide range of NO x (50 ppt to 30 ppb) were studied and include the Dickson, Tennessee, and Cornelia Fort Air Park sites during the 1999 Southern Oxidants Study, the Pellston, Michigan, site during the 1998 PROPHET summer intensive, and the Kejimkujik National Park site during the Atlantic 1996 study. Knowledge of NO x and HO x concentrations were critical for this evaluation. While NO x data are readily available at all sites, HO x data are limited. We employed a simple 10-reaction HO x model to calculate steady state OH radical concentrations as a function of [NO x ] to enable analysis of the data from all sites. Here, we use methyl vinyl ketone (MVK) concentrations to quantify the extent of isoprene-OH oxidation. Making use of the MVK/isoprene ratio, we show that the rate of production of isoprene oxidation products at various North American sites, although highly variable, exhibits the crossover from NO x -dependent to VOC-dependent conditions at $8 ppb [NO x ], in agreement with what is calculated from HO x measurements and our calculations.