Passive Sampler for Measurements of Atmospheric Nitric Acid Vapor (HNO3) Concentrations

Nitric acid (HNO3) vapor is an important nitrogenous air pollutant responsible for increasing saturation of forests with nitrogen and direct injury to plants. The USDA Forest Service and University of California researchers have developed a simple and inexpensive passive sampler for monitoring air concentrations of HNO3. Nitric acid is selectively absorbed on 47-mm Nylasorb nylon filters with no interference from particulate NO3-. Concentrations determined with the passive samplers closely corresponded with those measured with the co-located honeycomb annular denuder systems. The PVC protective caps of standardized dimensions protect nylon filters from rain and wind and allow for reliable measurements of ambient HNO3 concentrations. The described samplers have been successfully used in Sequoia National Park, the San Bernardino Mountains, and on Mammoth Mountain in California.


INTRODUCTION
Nitric acid (HNO 3 ) vapor, an air pollutant found in photochemical smog, affects the health of people and vegetation. Because of its very high reactivity and deposition velocity [1], HNO 3 provides large amounts of nitrogen (N) to forests and other ecosystems [2]. For example, in forests of the Los Angeles Basin mountain ranges, HNO 3 provides more that 60% of the dry-deposited N to mixed conifer forests [3]. Increased deposition of HNO 3 may lead to eutrophication of sensitive ecosystems and contamination of surface water with nitrate (NO 3 -), especially in locations close to photochemical smog source areas such as mountain ranges in southern California close to Los Angeles [4]. While ambient concentrations of HNO 3 expressed as 24-h averages did not exceed 2.1 µg/m 3 in a rural European location [5] and 4.8 µg/m 3 as a 12-h daytime average in the Sierra Nevada mountains in California [6], in the San Gabriel Mountains of southern California the 12-h daytime concentrations reached 28 µg/m 3 [7]. Information on HNO 3 spatial and temporal distribution in forest ecosystems of the western U.S. is essential for a better under-standing of its phytotoxic potential and development of N deposition models at a landscape level.
Measurements of HNO 3 concentrations using active sampling, such as annular denuder systems [8] or honeycomb denuder systems [9], although precise, are expensive and labor intensive. Use of these systems on a larger scale (landscape) is very difficult even if batteryoperated systems are utilized. Similarly to other pollutants, such as ozone, there is a clear need for simple and inexpensive device which would allow for landscape-level monitoring of the pollutant, including remote locations.
Although HNO 3 may be toxic to plants at elevated concentrations [10], its role as a source of N in different ecosystems is of primary importance. While for evaluation of toxic effects, peak concentration values of pollutants have to be known, for estimates of deposition, information on average concentrations of HNO 3 and its deposition velocity may be sufficient. Results from passive samplers, which integrate concentrations over a collection period, provide such information. Passive samplers, devices that are relatively inexpensive, low-maintenance, and do not require electric power, can be easily used for large-scale monitoring networks providing valuable information for the development of landscape-level pollutant distribution models.

EXPERIMENTAL METHODS
Nylon filters of 47 mm diameter (Nylasorb, Pall Corporation) have been used as a collection medium for HNO 3 . Such filters have been routinely used in annular denuder systems and filter packs for collection of this pollutant [11]. The filters were mounted in PVC double rings of 39 mm inner diameter (total surface of the exposed portion of a filter on both sides = 0.002389 m 2 ). Evaluation of filter responses to changing levels of HNO 3 , absorption capacity of filters, and performance of various protective caps was done in the continuously stirred tank reactors (CSTR) with controlled concentrations of the pollutant [12]. For testing the absorbing capacity of nylon filters, 30 of those mounted in the PVC rings were exposed to controlled high concentrations of HNO 3 in the CSTR chamber. Every 24 h, three replicate filters were removed from the chamber during 240 h of exposures. For calibration, additional exposures of nylon filters (PVC double rings without caps) were performed at low doses of the pollutant. Calibration curve was developed based on the combined CSTR results below the filter saturation point. Concentrations of HNO 3 and particulate NO 3 in CSTR chambers were monitored with honeycomb denuder/filter pack systems operating at 10 L/min flow rate, which provided 2.2 µm cutoff point for particulate matter [9]. Honeycomb denuder/filter pack systems were changed at the same time as passive samplers.
Preliminary tests indicated that, due to high reactivity of HNO 3 with surfaces, no pre-screens or diffusion tubes for controlling rate of air moving to sampling media could be used (no detectable amounts of the pollutant reached the nylon filters). Instead, HNO 3 was collected on filters well protected from rain and wind by PVC caps, and calibrated against the co-located honeycomb denuder systems. Tests of several designs of protective caps for nylon filters mounted in the PVC rings were performed in the CSTR chambers.
After exposures, nylon filters were placed in 250 ml Erlenmeyer flasks into which 20 mL of distilled, deionized water was added. The flasks were shaken on a wrist-action shaker for 15 min at middle speed. Amount of absorbed HNO 3 was determined as NO 3 with ion exchange chromatograph (Dionex Model 4000i). Results of the absorbed NO 3 were expressed as mg NO 3 -/m 2 of the total filter surface (two sides). Denuders and filters from honeycomb denuder/filter pack systems were extracted in 0.0024 M Na 2 CO 3 + 0.0030 M NaHCO 3 solution. Concentrations of NO 3 were also determined with ion exchange chromatography (Dionex Model 4000i).
Field tests of samplers were done in Sequoia National Park of the western Sierra Nevada, California. In summer 1999, passive samplers for HNO 3  Student's t test was used for comparison of performance of passive samplers against active monitors/honeycomb denuder systems. Differences are reported as significant at p < 0.05 [13].

RESULTS AND DISCUSSION
Deposition of HNO 3 to nylon filters (measured as NO 3 -) was proportional to the HNO 3 dose in the CSTR exposure chambers up to about 170 mg NO 3 -/m 2 (double-sided area of nylon filter). Above that level, nylon filters were not absorbing NO 3 linearly indicating near saturation point with HNO 3 (HNO 3 dose of about 5700 µg/m 3 × h). These results indicate that the open-ring filters would saturate after 14 days of exposure to average concentration of HNO 3 of about 17 µg/m 3 (6.6 ppb). A strong linear relationship between concentrations of the pollutant and its deposition to nylon filters allowed for precise calibration of passive samplers (Fig. 1). While absorption of NO 3 by nylon filters was closely correlated with HNO 3 ambient concentrations (R 2 = 0.9954), no relationship was found with the ambient concentrations of particulate NO 3 - (Fig. 2). The following formulas for calculating HNO 3 concentrations with passive samplers have been developed: where K values (slope of calibration curves) for various protective caps are listed in Table 1.   Protective caps reduced the rate of HNO 3 uptake by nylon filters and increased the K values (Table 1). These tests indicated that enclosure of filters in caps of various designs allows for HNO 3 collection at higher concentrations or longer collection periods.
Field tests in Sequoia National Park indicated that the highest HNO 3 concentrations determined both with passive samplers and honeycomb denuders were recorded at 504 m altitude at Ash Mountain, and the lowest concentrations were at 2,207 m altitude at Wolverton. Ranges of HNO 3 concentrations determined with honeycomb denuder systems and passive samplers were similar (0.6 -2.1 and 0.4 -2.9 µg/m 3 , respectively) (Fig. 3). Statistical evaluation of the HNO 3 data indicated that in four out of five sites where both honeycomb denuder systems and passive samplers were used, concentrations did not significantly differ. This suggests that these HNO 3 passive samplers could be used for estimates of ambient levels of the pollutant in remote mountain locations. However, due to different collection regimes (2-week-long integrated values for passive samplers versus two 24-h measurements per month during the growing season with honeycomb denuder systems), such a comparison has to be viewed with caution. These results also indicate that two different sampling regimes and monitoring methods provided very similar estimates of HNO 3 contamination in the study area. Precision of measurements with the HNO 3  passive samplers was good -the coefficient of variation of three replicate measurements stayed between 0.8 and 25.4% (average 9.2%) during the entire monitoring period. Precision of HNO 3 passive samplers could be additionally improved if a diffusion barrier providing uniform flow of air to the collection medium could be used.
A protective cap developed at the Department of Environmental Sciences of the University of California (UC-Riverside short cap) provided good protection of nylon filters against wind and rain (Fig. 4). This type of a protective cap has been successfully used in monitoring studies in remote mountain ranges of the San Bernardino and Mammoth Mountains. Results of these experiments are presented in other papers published in this volume [14,15].
In summary, it can be said that the developed sampler provides much-needed information regarding ambient levels of HNO 3 vapor on a large scale. However, in order to improve performance and reliability of the sampler, especially in conditions of high winds (a situation quite typical for these mountain environments), additional tests of diffusion barriers controlling flow of air to the collecting medium are needed. The USDA Forest Service scientists will perform such tests during the 2001 summer season.

1.
A strong relationship between ambient concentrations of HNO 3 and nitrate collected to nylon filters was determined in the CSTR chambers. The nylon filters used in this study (effective double-sided area of 0.002389 m 2 ) have a capacity of collecting 5,700 µg HNO 3 /m 3 × h which allows for long-term (several weeks) monitoring of the pollutant. Protective caps of various designs reduce rate of HNO 3 deposition and amount of nitrate absorbed on the collecting medium and extend potential time of HNO 3 collection. 2. Field tests demonstrated that results obtained with passive samplers were similar to the values recorded with honeycomb annular denuder/filter pack systems. Therefore, we conclude that passive samplers can reliably measure ambient levels of HNO 3 . 3. Low price, simple design, and easy operation of the samplers allow for a large (landscape) level monitoring of HNO 3 . This offers a possibility for evaluation of risks to ecosystems and humans. 4. Additional tests are needed for improving performance of the samplers. Enclosing nylon filters in collection devices providing steady airflow by diffusion barriers will be tested.