Long-Term Measurements of NO3 Radical at a ... - ACS Publications

Nov 9, 2009 - Nitrate radical (NO3), an important nighttime tropospheric oxidant, was measured continuously for two years (July 2005 to September 2007...
0 downloads 0 Views 1MB Size
Download PDF
Environ. Sci. Technol. 2009 43, 9117–9123

Long-Term Measurements of NO3 Radical at a Semiarid Urban Site: 1. Extreme Concentration Events and Their Oxidation Capacity D A V I D A S A F , * ,† D A N I E L P E D E R S E N , † VALERI MATVEEV,† MORDECHAI PELEG,† C H R I S T O P H K E R N , ‡ J U T T A Z I N G L E R , ‡,§ ULRICH PLATT,‡ AND MENACHEM LURIA† Institute of Earth Sciences, Hebrew University of Jerusalem, Givat Ram, Israel, and Institute of Environmental Physics, University of Heidelberg, Heidelberg, Germany

Received March 17, 2009. Revised manuscript received October 4, 2009. Accepted October 28, 2009.

Nitrate radical (NO3), an important nighttime tropospheric oxidant, was measured continuously for two years (July 2005 to September 2007) in Jerusalem, a semiarid urban site, by longpath differential optical absorption spectroscopy (LPDOAS). From this period, 21 days with the highest concentrations of nitrate radical (above 220 pptv) were selected for analysis. Joint measurements with the University of Heidelberg’s LPDOAS showed good agreement (r ) 0.94). For all daytime measurements, NO3 remained below the detection limit (8.5 pptv). The highest value recorded was more than 800 pptv (July 27, 2007), twice the maximum level reported previously. For this subset of measurements, mean maximum values for the extreme events were 345 pptv (SD ) 135 pptv). Concentrations rose above detection limits at sunset, peaked between midnight and early morning, and returned to zero at sunrise. These elevated concentrations of NO3 were a consequence of several factors, including an increase in ozone concentrations parallel to a substantial decrease in relative humidity during the night; Mean nighttime NO2 levels above 10 ppbv, which prevented a deficiency in NO3 precursors; Negligible NO levels during the night; and a substantial decrease in the loss processes, which led to a lower degradation frequency and allowed NO3 lifetimes to build up to a maximum mean of 25 min. The results indicate that the major sink pathway for NO3 was direct homogeneous gas phase reactions with VOC, and a smaller indirect pathway via hydrolysis of N2O5. The Jerusalem measurements were used to estimate the oxidation potential of extreme NO3 levels at an urban location. The 24 h average potential of NO3, OH, and O3 to oxidize hydrocarbons was evaluated for 30 separate VOCs. NO3 was found to be responsible for approximately 70% of the oxidation of total VOCs and nearly 75% of the olefinic VOCs; which was more than twice the VOC oxidation potential of the OH radical. These

* Corresponding author phone: +972-2-6584827; fax: -+972-25637260; e-mail: [email protected]. † Hebrew University of Jerusalem. ‡ University of Heidelberg. § Current address: Institute for Meteorology and Climate Research, Karlsruhe Institute of Technology, Germany. 10.1021/es900798b CCC: $40.75

Published on Web 11/09/2009

 2009 American Chemical Society

results establish the NO3 radical as an important atmospheric oxidant in Jerusalem.

1. Introduction The nitrate radical (NO3) is a key component of nighttime oxidation chemistry in the troposphere, often comparable to the OH radical as a sink for nitrogen oxides and VOCs (1-3). It is therefore considered a major atmospheric oxidant for species with high reactivity toward it. Atmospheric NO3 chemistry has been discussed extensively elsewhere (2, 4); The Supporting Information (SI) includes rate coefficients and details of VOC reactions. NO3 is formed in the boundary layer when NO2 reacts with O3 (R1), and its production rate, PNO3, can be calculated by (R2) (1). In the present study, typical nighttime concentrations of NO2 (15 ppbv) and ozone (40 ppbv) at 298 K, produced an estimated 0.5 pptv of NO3 per second. NO2 + O3 f NO3 + O2

(R1)

PNO3 ) k1[NO2][O3]

(R2)

Photolysis is an important daytime loss mechanism of NO3, with a photolytic lifetime of ∼5 s (4). Recently, however, daytime mixing ratios of NO3 of up to 30 ppt have been reported close to sunset (5). The major gas phase sink in NOx-rich environments is the rapid reaction with NO (R3), which occurs day and night, and may limit NO3 lifetime near NO sources (6). At nighttime, however, NO is rapidly oxidized by ozone (R4) and its depletion may allow NO3 to accumulate. NO3 + NO f NO2 + NO2

(R3)

NO + O3 f NO2 + NO2

(R4)

In the continental boundary layer, NO3 can also be directly removed by reactions with VOCs, although most of them react much more slowly with NO3 than with OH (6). Nevertheless, NO3 is very important because it is usually present at night at mixing ratios much higher than OH (up to 800 pptv in the present study), and reaction rates of some hydrocarbons with NO3 and OH are comparable (7). Another major scavenging process beside direct NO3 sinks is an indirect pathway with a fast, temperature dependent equilibrium to produce N2O5 in the gas phase (R5) (ref 8, also SI). Part of the N2O5 decomposes thermally to produce NO3 and NO2 again, but subsequent removal of N2O5 by reaction with water to form HNO3 (R6) leads to a net loss of NO3 from the atmosphere. This hydrolysis can be homogeneous or heterogeneous (on particles) (1, 9). Heterogeneous removal can be the most important in the nighttime scavenging processes for NO3, especially during winter (e.g., refs 1, 4, 9, 10). M

NO3 + NO2 798 N2O5

(R5)

N2O5 + H2O f 2HNO3

(R6)

This paper describes extreme episodes of NO3 measured in Jerusalem, Israel, an urban semiarid area, and reports outstanding high mixing ratios of up to 807 ppt NO3, 2-fold greater than measured anywhere else. NO3 concentrations are presented and analyzed for twenty one different days selected from a two year campaign, with the highest mixing ratios of nitrate radical, and maximum nighttime means VOL. 43, NO. 24, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

9117

FIGURE 1. A time series (local time GMT+2) for the mixing ratios of NO3, NO2, O3, and RH during the day with the highest levels of NO3 ever reported. Sunrise and sunset are marked with dashed lines. Error bars for NO3 (usually smaller than the markers) represent standard uncertainties, which incorporate the integration of spectra, instrumental and atmospheric noise, the path length, and uncertainties in the integrated differential cross section of NO3. greater than 220 ppt. The unique chemistry of these extreme levels is presented, along with their effect on oxidation capacities of VOCs in a typical urban environment. Additionally, we evaluated the degradation frequencies of NO3 and the contribution of these elevated concentrations to the oxidation efficiency of VOCs in the atmosphere, compared with the contributions of ozone and the OH radical.

2. Experimental Details and Methods Additional technical details on DOAS instruments, detection limits, and data processing are available in the SI. 2.1. Experimental Setup. The NO3 measurements were performed in Jerusalem (31°47′N 35°13′E), using long-path differential optical absorption spectroscopy (LP-DOAS) (11), the same instrument used for NO3 measurements described previously (11-13). The 3.8 km light path passed over the mountainous terrain of Jerusalem (SI Figure S2), and the height of the DOAS beam ranged between 15 and 100 m above ground level. The beam passed over an urban area containing anthropogenic emission sources, primarily from traffic, with roads from all directions, including a major six lane highway. During a continuous two year campaign (July 2005 to September 2007), extreme mixing ratios of NO3 (above 220 ppt) were observed on 21 separate days. The entire campaign will be discussed in a later publication. NO2 and NO3 were measured by the DOAS, representing average concentrations in the air mass traversed by the light beam. O3, NO/NOx, and meteorological parameters were measured continuously at the HUJI laboratory site using conventional analyzers, representing point parameters at the measuring site. Parallel DOAS measurements were conducted at the Hebrew University site (July 26 to August 25, 2005), with Heidelberg University’s high end LP-DOAS (HEIDDOAS), which used retro-reflectors positioned 4.8 km from the shared base station, providing an effective light path of 9.6km (SI Figures S1, S2). The objectives included a comparison with the HUJI-DOAS. 2.2. Raw Data Processing. NO3 concentrations were obtained from measured spectra using evaluation procedures described previously (13), and briefly summarized here. Background spectra, dark current, and electronic offset were subtracted, followed by band-pass filtration of the resulting spectra. A fifth order polynomial and reference spectra for 9118

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 24, 2009

NO2 (measured on site), NO3 (literature values), and water vapor (from daytime spectra with no NO3 present) were fitted using nonlinear least-squares fitting routines (10) by the analysis software MFC (14). Finally, the concentrations calculated from the fit results were adjusted for pressure and temperature. NO2 and NO3 were measured through absorbance spectra in the 418-442 and 650-674 nm range, respectively. In this campaign, NO3 was quantified at 662 nm.

3. Results and Discussion 3.1. Record High NO3 Levels (25-26 July, 2007). The highest NO3 mixing ratios of the campaign were observed on July 25-26, 2007 (Figure 1). NO2, O3, RH and the nighttime period (solar radiation less than 0.05 kWatt/M2) are also presented. The detection limit of the DOAS for NO3 was below 5 ppt; hence possible concentrations of NO3 during daytime could not be resolved. NO3 levels began to increase at sunset and built up to a temporary maximum of ∼200 pptv between 9 and 11 pm, followed by a decrease to below 30 ppt just before midnight. After this time, NO3 levels rose above 200 ppt between 1 and 5 am, peaking at 807 pptv at 4 am, the highest value recorded during campaign. This is a factor of 2 greater than measurements reported by Platt (11) in Riverside, California, and the Houston, Texas aircraft measurements performed by Brown (15). This extreme buildup of NO3 corresponded with several phenomena: RH decreased from 60 to 30%, ozone rose from 10 ppb up to 60 ppb, and NO2 rose from 5 to 20 ppb. A similar pattern was observed on all days with extreme levels of NO3. 3.2. Extreme NO3 Radical Levels. 3.2.1. Mixing Ratios of NO3. During the two year campaign (July 2005 to September 2007), the highest mixing ratios of NO3 (>220 pptv) were observed on twenty one separate days (SI Figure S3), which were selected for analysis of the nocturnal hours (18:00-08: 00) presented here. Mixing ratios of NO3 rose above the detection limit at sunset, peaked between midnight and early morning, and decreased to zero at sunrise. The mean maximum value for the nighttime NO3 was 345 pptv (SD ) 135 pptv). The highest NO3 levels were generally observed during warmer months, most commonly in July and August.

FIGURE 2. A time series comparison of the joint field study between the HEID and HUJI measurements during two episodes, 2-6 and 11-17 of August, 2005. Notice different y scales for the two plots. Nevertheless, elevated NO3 levels occurred in all four seasons, albeit with lowest frequency in winter (SI Figure S3). 3.2.2. Meteorology. Time profile plots of hourly mean values for temperature and relative humidity (RH) show that the median temperatures during nighttime were relatively warm at 295-300K, with no significant variations. RH, on the other hand, demonstrated a unique pattern: a rise after sunset up to 55%, followed by a substantial decrease after 22:00 to 35% (SI Figure S4a). The mean wind speed was 2.5 m/s, and the strongest winds and predominant wind direction prevailed from the West and North-West quadrants (SI Figure S4b). 3.2.3. Quality Control. NO3 concentrations measured by the two DOAS systems in the joint study were compared for quality control (Figures 1 and SI S5). Although the two light paths began at the same point, and were fairly close to each other within the NBL, they represent different air masses and passed over different terrain. The HEID-DOAS light path passed at a low altitude over a six lane highway in the first 1km (SI Figures S1 and S2), and was therefore exposed to significant NO concentrations. The HUJI-DOAS light path, in contrast, passed far above regular city streets and intersections, and was further away from NO emission sources. It is reasonable to assume that the HEID-DOAS was more influenced by local NO emissions, which would eliminate NO3 by reaction (R3), and that the two instruments might have sampled different air masses, due to strong inversion and stability. Despite the difference in the air masses measured by each of the DOAS instruments, a close agreement was observed between the data sets. The pattern and timing of the variations in NO3 concentrations was very similar (Figure 2). The magnitude of the measurements showed a systematic trend, with peak concentrations for HEID one-third lower than for HUJI, as expected, if the HEID-DOAS was exposed to higher levels of NO. For both instruments, NO3 levels remained below DL between August 6-11 (data not presented). The

Pearson correlation coefficient between the two data sets was a highly significant r ) 0.94 (SI Figure S5). 3.2.4. Mixing Ratios of O3, NO2, and NO. Median hourly values of NO3, NO2, O3, and NO are shown for all twenty one days (Figure 3). Significant variations can be seen during the nighttime for all species (More details can be seen in Figure 1). Hourly average NO3 mixing ratios showed a rise from sunset until 22:00, when the optimal conditions for its production appeared (as described later), and then remained stable at 60-80 pptv until 01:00. After that, NO3 increased again to a mean of 160-200 pptv between 01:30 and 06:00. NO2 hourly mean levels decreased gradually from sunset until 01:00 a.m., then remained stable until sunrise, when they climbed parallel with the morning rush hour. The NO3 profile did not follow the NO2 profile, in contrast with previous studies (9), which reported similarities in the seasonal and diurnal profiles of NO2 and NO3. Due to the abundance in NO2 in Jerusalem, there was no deficiency in NO3 precursors, and NO3 profiles were, therefore, not affected directly by variations in NO2 (Figure 3). Nocturnal profiles of O3, unlike NO2, were more similar to NO3 between 22:00 and 06:00, rising from average levels of 25 to 40 ppb. This ozone pattern has frequently been observed in Jerusalem, as well as in other cities in Israel (data from the National Air Monitoring Network of the Israel Ministry of Environmental Protection, (16)). O3 is not produced photochemically at night, and its nighttime rise could not be explained by horizontal transport of O3, both because of the very weak and variable winds at night and because the changes in O3 mixing ratios occur simultaneously in other cities. Two reasons have been suggested for the nocturnal ozone peak: The most likely explanation is the possibility of a persisting low-level inversion trapping the daytime ozone. With the gradual lowering of the inversion layer during nighttime, the ozone level increased, peaking between 03:00 and 05:00. Another reason might be an intrusion event during which dry air from the free troposphere subsided into the nocturnal boundary layer, as suggested by Steinberger (16). Most of the time, NO appeared to have no influence on NO3 along the light path, even though NO point measurements reached 10 ppbv during specific periods at night, which would be expected to affect NO3 according to reaction (R3). These values of NO should yield a very short NO3 lifetime, less than 1 s, and drastically lower the NO3 mixing ratios to zero levels, but that was not the case. The probable explanation is that the NO measured at the sampling station was elevated compared to average levels in the boundary layer as a whole, since the station is located much closer to NO emission sources that the HUJI-DOAS light path, which passes high above the roads (SI Figure S2). Therefore, NO would not be expected to be well mixed in the stable nocturnal boundary layer, since it is rapidly consumed by its reaction with O3, allowing NO3 to build up and maintain higher concentrations, as measured by the DOAS in the boundary layer. In addition, a decrease of NO to almost zero levels can be seen from 01:30 until sunrise, allowing vertically transported O3 to build up and produce additional NO3. Nevertheless, the importance of NO for the removal of NO3 should not be neglected in an urban location with local emission sources. The 21 episodes of maximum NO3 mixing ratios were thus associated with an intrusion of ozone-rich air that occurred parallel to decreases in NO2 and RH, an average of 2 h before the second rise in NO3 to unusually high mixing ratios between 200 and 800 ppt. The rise in O3 would be expected to cause an immediate increase in NO3, but as explained in detail in section 3.3, the most probable explanation for the delay is the depletion of NO3 by reactions VOL. 43, NO. 24, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

9119

FIGURE 3. Hourly means of the mixing ratios ((1 SD) of (a) NO3, (b) NO2, (c) O3, and (d) NO for the days with extreme NO3 events during HUJI campaign. with VOC. The increase in NO3 levels probably occurred only after these VOC were exhausted. 3.2.5. NO3 Production and Degradation Rates. The mean rate of production of NO3 (PNO3) ranged between 0.2 and 0.45 pptv s-1, and remained stable throughout the night (SI Figure S6a). Maximum PNO3 was 1.7 pptv s-1, due to the elevated concentrations of NO2 and O3, thus producing extremely high NO3 concentrations. A linear regression of PNO3 vs NO3, PNO3 ) 137.15[NO3] + 79.9, (R2 ) 0.24), indicates a possible contribution of the direct sinks. Heintz (2) and others suggested that PNO3 is predicted to be proportional to [NO3] if first order losses dominate. The mean total degradation frequency of NO3 (fNO3) ranged from 1 × 10-4 s-1 to almost 0.02 s-1 and showed a reverse trend to increases in NO3, as expected (SI Figure S6b). 3.2.6. NO3 Lifetime. The lifetime of NO3 is a useful diagnostic tool for analyzing field observations. This can be estimated by assuming a local stationary state between all the loss processes of nitrate radical and its formation by reaction (R1), and then calculating the effective first-order rate constant, f(NO3), for loss of NO3. The e-lifetime of nitrate radical is then given by τNO3 ) 1/f(NO3). The half-hourly mean of τNO3 ranged between 1 and 25 min, with an average between sunset and sunrise of 12.5 min (SI Figure S6c). It exhibited the same temporal pattern as NO3 mixing ratios: a moderate rise until 22:00, then remaining stable up to 01:00 a.m., followed by another rise until 03:00, which remained stable until sunrise. These analyses included only those data which were estimated to be under steady state, and therefore all measurements from sunrise and sunset, and those with variability in NO2 and NO3, were excluded. This suggests that an approximate steady state, where the production and loss were balanced, might possibly have been achieved temporarily during the periods in which NO3 lifetime and concentrations remained stable. In addition, Brown (17) 9120

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 24, 2009

suggested that the time to steady state is strongly temperature dependent, and demonstrated through observations supported by model simulations that a steady state approximation can be reached in similar conditions to this study. Furthermore, for the warmer temperatures typical of this study, steady state occurs rapidly. A thermal equilibrium was probably achieved for (R5), since thermal N2O5 lifetime is on the order of 20 s, whereas the average NO3 lifetime is on the order of 20 min (1200 s). Also, from Keq and NO2 levels of ∼20 ppb it follows that [N2O5]/ [NO3] is around 10. Thus, if indirect losses dominate, N2O5 lifetime would be about 10× the observed NO3 lifetime, i.e., around 12 000 s. This is 3 orders of magnitude longer than the thermal lifetime of N2O5, so thermal equilibrium is a very good approximation. 3.3. Contribution of the Different Sinks of NO3. The relative importance of direct losses (via VOCs) vs indirect losses (via N2O5) for Jerusalem, were evaluated using two analysis methods. One method (1, 2, 4, 18, 19) used the inverse dependence between τ(NO3) and [NO2] to qualitatively assess the importance of heterogeneous hydrolysis of N2O5. Brown (17) presented a slightly different analysis, which explicitly accounts for the temperature dependence of Keq, and is derived from the expression for the steady state lifetime of NO3. τ(NO3) ≡

[NO3] ≈ (kNO3 + KN2O5Keq(T)[NO2])-1 k1[O3][NO2] (E1)

Equation E1 suggests that a plot of τ(NO3)-1 against Keq(T)[NO2] should give a straight line whose slope and intercept are the effective first-order loss rate coefficients for N2O5 (kN2O5) and NO3 (kNO3), respectively.

A linear regression of the first analysis, ln(τ[NO3]) vs ln([NO2]) for the entire data, yielded the relationshipln (τ[NO3]) ) -0.85 × ln[NO2] + 7.95, R2 ) 0.45). (SI Figure S7a). The slope of -0.85 implies that there was a considerable contribution of indirect NO3 loss, although, since the slope is not -1, indirect removal of NO3 was not the exclusive pathway. The observation from the first analysis suggests that under conditions of high water concentrations, hydrolysis, either heterogeneous or possibly homogeneous, is an important mechanism. Since the hygroscopic growth of aerosols increases the surface available for heterogeneous reactions, the N2O5 uptake on aerosols could contribute substantially to the indirect removal of NO3 and the conversion of NOx into HNO3 (20). Similar observations were seen in previous studies such as Heintz for Kap Arkona (2) and Verkoussis for the Minos campaign (9). In the Jerusalem region, long-term trends of aerosol levels, especially sulfate, are elevated and have a relatively constant composition (21). In addition, no dust storm events occurred during elevated NO3 events, and back trajectories showed that all of the air masses originated from eastern Europe, thus supporting the conclusion that aerosol composition did not vary substantially, since regional aerosol content is mainly affected by long-range transport (21). As a result of these considerations, gamma variability is expected to be very low, therefore decreasing any possible confounding by RH and water content. Nighttime RH decreases allowed NO3 to build up to high levels, with NO3 levels increasing almost by a factor of 3 when the RH decreased below 60%. This negative relationship is seen even more clearly between NO3 and H2O mixing ratios (SI Figure S8a), and between the lifetime of NO3 and H2O mixing ratios (SI Figure S8b). The second analysis, as stated above, incorporated the temperature dependence of Keq into eq E1 (15, 17), and yielded clear evidence for a rapid direct sink for NO3, kNO3-1 ) 6.6 min and kN2O5-1 ) 175 min (SI Figure S7b). Degradation frequencies for N2O5 are converted into their corresponding frequencies for NO3 using eq E2: findir ≡ Keq(T)[NO2]fN2O5

(E2)

Thus the effect of the indirect loss contributes to a lifetime k-1 of about 17 min, which is 1/3 of the direct loss. NO3 lifetimes calculated using only the degradation of NO3 by VOCs (see section 3.4), yielded lifetimes of approximately ∼600 s, similar to the average NO3 lifetimes of ∼750 s calculated from the steady state assumption in eq 1 between 21:00 and 01:00. This demonstrates the importance of VOCs as a sink for NO3. It seems probable that NO3 removal by VOCs dominated until 01:00 (corresponding with τ(NO3) of ∼600 s), followed by the titration of VOC from the boundary layer, as indicated by the rise in τ(NO3) up to an average of 1500 s, allowing NO3 to accumulate. Both methods of analysis used to evaluate the removal of NO3 support the conclusion that the direct removal of NO3 dominates (estimated at about 60%), most likely by reactions with VOCs, although the indirect removal of NO3 is also significant, at approximately 30%, and cannot be neglected (SI Figure S7b). 3.4. Oxidation Potential. Daytime oxidation of hydrocarbons proceeds mainly by reactions with OH and O3. For nighttime chemistry, NO3 plays an important role due to its high reactivity and the negligible amount of OH at night (7). The oxidation capacities of NO3, O3, and OH with individual selected VOCs were compared in order to estimate the contribution of NO3 to the atmospheric oxidation capacity when present at extreme levels. NO3 is highly reactive with unsaturated VOCs in general, and with certain anthropogenic compounds in particular,

e.g., phenol and cresols emitted from traffic (e.g., ref 7). However, under urban conditions, NO, which is present at elevated levels, is probably the most important NO3 scavenger, and often neither natural nor anthropogenic VOCs play an important role (e.g., 19). OH mixing ratios were not measured in this campaign, and were estimated using a one-dimensional chemical transport model, UAHCTM_1D (22). This model includes an explicit gas phase chemical mechanism and takes into account the vertical motion of the different species based on diffusion and advection calculations and on deposition velocity values. The basic photochemical processes described by 166 gas-phase reactions were based on the Trainer mechanism (23), which was updated with relevant NO3 chemistry reactions according to Atkinson (24). A detailed description of the model and its mechanism is given by Tas (25). The modeled OH concentrations were then confirmed by a comparison to OH levels reported at similar urban locations (26-29). From a group of 30 major VOCs including alkanes, alkenes, alkynes, and aromatics, the oxidation capacities of the radicals X () NO3, OH, O3) were calculated. Rate constants of the selected VOCs with these oxidants were taken from Atkinson (6, 24). Unfortunately, VOC measurements were not conducted during the campaign; therefore, typical mixing ratios for the 30 different VOCs representative of urban locations were estimated and averaged from several campaigns reported in the literature (SI Table S1). The VOC concentrations chosen for this calculation are intended to represent conditions typical of an urban location, and not the actual conditions in Jerusalem. The 24 h integrated oxidation capacity was calculated via (E3) for each of the oxidants as suggested by Geyer (1). OC )

∑k

HCi-X[HCi][X]

(E3)

i)1

Since the NO3 nighttime levels were so high, and the estimated daytime concentrations were negligible in comparison, they did not influence the 24 h integral of the nitrate radical oxidation capacity. The nighttime oxidation capacity of the nitrate radical ranged between (1.1 and 90) × 107. A 24 h integral of the NO3 concentration for all the days was calculated, yielding an average night-time oxidation capacity of 13.1 × 107 cm-3s-1. Ozone exhibited a much smaller value of 6.5 × 106 cm-3s-1, integrated over 24 h, with a maximum oxidation capacity value of 1.1 × 107 cm-3s-1. The 24 h mean of the OH oxidation capacity was 5.54 × 107 cm-3s-1 (less than half that of the NO3), with a maximum value of 17 × 107 cm-3s-1. These values of the oxidation capacities demonstrate the huge potential of the NO3 radical as an oxidizing agent at these high mixing ratios, more than twice as high as the OH radical. The relative contribution of NO3 to the oxidation of the VOCs was found to be 68%, vs 29% for OH (SI Figure S9a), even though the rate constant for NO3 reactions with most alkanes and aromatic compounds was much smaller than for OH. This can be explained by the elevated mixing ratios of the nitrate radical, as measured in Jerusalem. These results were sensitive to the concentration of VOCs which were highly reactive with NO3, e.g., cresol, which has a rate constant for NO3 similar to OH. When adding concentrations of cresol at typical urban concentrations (SI Table S1), the relative contribution of NO3 increased dramatically to 88%, compared to 11% for OH and 1% for ozone (SI Figure S9b). These calculations demonstrate that, when evaluating the oxidation capacity of NO3, it is important to account carefully for concentrations of species that are highly reactive with NO3, and to include them specifically in any measurements or calculations. VOL. 43, NO. 24, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

9121

TABLE 1. Comparison of the NO3 and Oxidation Potential in the Boundary Layer average values nighttime average maxima of NO3, ppt maximum average of τ(NO3), sec relative oxidation capacity of VOCs relative oxidation capacity of olefinic VOCs reference

(2)

Pabstthum (summer)

Jerusalem (all year)

(3)

∼10, peak of 70

∼200, peak of 800

(4)

∼100-400 O3, 17% NO3, 28% OH, 55% O3, 19% NO3, 31% OH, 50% 1

∼100-1500 O3, 3% NO3, 68% OH, 29% O3, 4% NO3, 73% OH, 23% this study

(5)

(6) (7)

Major compounds like isoprene and monoterepenes, which are emitted largely by vegetation, react more readily with NO3, but have a very small contribution to the total VOCs in an urban location, approximately 1% for these calculations. Olefinic compounds such as alkenes, subsets of the VOCs, have a greater reaction rate with NO3, and were therefore also considered separately (SI Figure S9c). As expected, the contribution of NO3 to the oxidation of olefin compounds increased to nearly 75%, while the hydroxyl radical decreased below 25% (Table 1). The results of oxidation potential of this campaign were compared with observations in Pabstthum (52°51′15′′N 12°56′25′′E), a rural village in Germany (1). A large difference is observed both for the NO3 mixing ratios, and for the oxidation potentials (Table 1). Both mixing ratios and lifetimes of NO3 were significantly higher in Jerusalem, reflecting the differences in the characteristics of the sites, rural vs urban. As a result, the relative oxidation capacity of NO3 for VOCs at Jerusalem was more than 2-fold higher than in Pabstthum, where it contributed less than one-third of the VOC removal. This demonstrates that NO3 can be the dominant oxidant for VOCs in an urban location, especially when present at elevated concentrations. In addition, the NO3 oxidation potential could be even greater for olefinic VOCs, because of their higher affinity for reactions with NO3. These results provide strong evidence that NO3 can be the dominant contributor to VOC degradation, and establish the importance of the NO3 radical as an atmospheric oxidant in Jerusalem.

(8)

(9)

(10)

(11) (12) (13) (14) (15)

Acknowledgments This project was funded by the German-Israeli Foundation for Scientific Research and Development (GIF), grant I-776288/2003. We are grateful to the Gilo Community center for use of their rooftop for the full 2 years of the project, and Eretz Hatzvi Elementary School for use of their rooftop during the pilot project in 2005. Supporting air quality data was provided courtesy of the Israel Ministry of the Environment.

Note Added after ASAP Publication The captions for Figures 1 and 2 appeared incorrectly in the version of this paper published ASAP November 9, 2009; the corrected version published ASAP November 13, 2009.

Supporting Information Available

(16) (17)

(18) (19) (20) (21)

Further experimental details, figures, and table. This material is available free of charge via the Internet at http:// pubs.acs.org.

(22)

Literature Cited

(23)

(1) Geyer, A.; Alicke, B.; Konrad, S.; Schmitz, T.; Stutz, J.; Platt, U. Chemistry and oxidation capacity of the nitrate radical in the 9122

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 24, 2009

continental boundary layer near Berlin. J. Geophys. Res. 2001a, 106, 8013–8025. Heintz, F.; Flentje, H.; Dubois, R.; Platt, U. Long-term observation of nitrate radicals at the Tor Station, Kap Arkona (Rugen). J. Geophys. Res. 1996, 101, 22,891–22,910. Smith, N.; Plane, J. M. C.; Nien, C. F.; Solomon, P. A. Nighttime radical chemistry in the San Joaquin Valley. Atmos. Environ. 1995, 29, 2887–2897. Geyer, A.; Ackermann, R.; Dubois, R.; Lohrmann, B.; Mueller, T.; Platt, U. Long-term observation of nitrate radicals in the continental boundary layer near Berlin. Atmos. Environ. 2001b, 35, 3619–3631. Geyer, A.; Hofzumahaus, A.; Holland, F.; Konrad, S.; Klupfel, T.; Patz, H. W.; Perner, D.; Schafer, H. J.; Volz-Thomas, A.; Platt, U. Nighttime production of peroxy and hydroxyl radicals during the BERLIOZ campaign: Observations and modeling studies. J. Geophys. Res. 2003, 108, 8249–8256. Atkinson, R. Atmospheric chemistry of VOCs and NOx. Atmos. Environ. 2000, 189/190, 431–435. Kurtenbach, R.; Ackermann, R.; Becker, K. H.; Geyer, A.; Gomes, J. A. G.; Lorzer, J. C.; Platt, U.; Wiesen, P. Verification of the contribution of vehicular traffic to the total NMVOC emissions in Germany and the importance of the NO3 chemistry in the city air. J. Atmos. Chem. 2002, 42, 395–411. DeMore, W. B.; Sander S. P.; Golden D. M.; Hampson R. F.; Kurylo M. J.; Howard, C. J.; Ravishankara, A. R.; Kolb, C. E.; Molina, M. J. Chemical Kinetics and Photochemical Data for Use in Stratospheric Modeling, JPL Pub. Evaluation No. 12; Jet Propulsion Laboratory: Pasadena, CA, 1997. Vrekoussis, M.; Kanakidou, M.; Mihalopoulos, N.; Crutzen, P. J.; Lelieveld, J.; Perner, D.; Berresheim, H.; Baboukas, E. Role of the NO3 radicals in oxidation processes in the eastern Mediterranean troposphere during the MINOS campaign. Atmos. Chem. Phys. 2004, 4, 169–182. Stutz, J.; Platt, U. Numerical analysis and estimation of the statistical error of differential optical absorption spectroscopy measurements with least square methods. J. Appl. Opt. 1996, 30, 6041–6053. Platt, U.; Perner, D.; Potz, H. Simultaneous measurement of atmospheric CH2O, O3 and NO2 by differential optical absorption. J. Geophys. Res 1979, 84, 6329–6335. Platt, U. Differential optical absorption spectroscopy (DOAS). In Air Monitoring by Spectroscopic Techniques, Chem. Anal. Servol. 127; Sigrist, M. W., Ed.; Wiley: New York, 1994; pp 24-87. Matveev, V.; Luria, M.; Alper-Siman Tov, D.; Peleg, M. Long range transportation of air pollutants from Europe towards Israel. Isr. J. Earth. Sci. 2002, 51, 17–28. Gomer, T.; Brauers, T.; Heintz, F.; Stutz, J.; Platt, U. MFC User Manual V. 1.98, Unst. Fur Umweltphysik der Univ: Heidelberg, Germany, 1993. Brown, S. S.; Dube, W. P.; Fuchs, H.; Ryerson, T. B.; Wollny, A. G.; Brock, C. A.; Bahreini, R.; Middlebrook, A. M.; Neumna, J. A.; Atlas, E.; Roberts, J. M.; Osthoff, H. D.; Trainer, M.; Fehsenfeld, F. C.; Ravishnkara, A. R. Reactive uptake coefficients for N2O5 determined from aircraft measurements during the second Texas Air Quality study: Comparison to current model parameterizations. J. Geophys. Res. 2009, 114, D00F10. Steinberger, E. H.; Ganor, E. High ozone concentrations at night in Jerusalem and Tel-Aviv. Atmos. Environ. 1980, 14 (2), 221–225. Brown, S. S.; Stark, H.; Ravishankara, A. R. Applicability of the steady state approximation to the interpretation of atmospheric observations of NO3 and N2O5. J. Geophys. Res. 2003, 108, 4539– 4546. Martinez, M.; Perner, D.; Hackenthal, E.; Kultzer, S.; Schultz, L. NO3 at Helgoland during the NORDEX campaign in October 1996. J. Geophys. Res. 2000, 105 (18), 22685–22695. Platt, U.; Janssen, C. Observations and role of free radicals NO3, ClO, BrO, and IO in the troposphere. Faraday Discuss. 1995, 100, 175–198. Carslaw, N.; Plane, J. M. C.; Coe, H.; Cuevas, E. Observations of the nitrate radical in the free troposphere at Izana de Tenerife. J. Geophys. Res 1997, 102, 10613–10622. Foner, H. A.; Ganor, E. The chemical and mineralogical composition of some urban atmospheric aerosols in Israel. Atmos. Environ. 1992, 26B, 125–133. Biazar, A.-P. The role of natural nitrogen oxides in ozone production in the southern environment. Dissertation, The University of Alabama in Huntsville, Hunstille, Al, 1995. Trainer, M.; Williams, E. J.; Parish, D. D.; Buhr, M. P.; Allwine, E. J.; Westberg, H. H.; Fehsenfeld, F. C.; Liu, S. C. Models and observations of the impact of natural hydrocarbons on rural ozone. Nature 1987, 329, 705–707.

(24) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Crowley, J. N.; Hampson, R. F.; Kerr, J. A.; Rossi, M. J.; Troe, J. Summary of evaluated kinetic and photochemical data for atmospheric chemistry: Volume II - gas phase reactions of organic species. Atmos. Chem. Phys. 2006, 6, 3625–4055. (25) Tas, E.; Peleg, M.; Pedersen, D. U.; Matveev, V.; Biazar, A. P.; Luria, M. Measurement-based modeling of bromine chemistry in the boundary layer: 1. Bromine chemistry at the dead sea. Atmos. Chem. Phys. 2006, 6, 5589–5604. (26) George, L. A.; Hard, T. M.; O’Brien, R. J. Measurement of free radicals OH and HO2 in Los Angeles Smog. J. Geophys. Res. 1999, 104, 11643–11655.

(27) Hard, T. M.; Chan, C. Y.; Mehrabzadeh, A. A.; Pan, W. H.; O’Brien, R. J. Diurnal cycle of tropospheric OH. Nature 1986, 332, 617–620. (28) Hard, T. M.; Chan, C. Y.; Mehrabzadeh, A. A.; O’Brien, R. J. Diurnal HO2 cycles at clean air and urban sites in the troposphere. J. Geophys. Res. 1992, 97, 9785–9794. (29) Ren, X.; Harder, H.; Martinez, M.; Lesher, R. L.; Oliger, A.; Shirley, T.; Adams, J.; Simpas, J. B.; Brune, W. H. HOx concentrations and OH reactivity observations in New York City during PMTACS-NY2001. Atmos. Environ. 2003, 37, 3627–3637.

ES900798B

VOL. 43, NO. 24, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

9123