Reactivity and Loss Mechanisms of NO₃ and N₂O₅ at a Rural Site on the North China Plain

Abstract

NO₃, NO₂, O₃, and relevant parameters were measured at a rural site on the North China Plain during June 2014. During the campaign, the average concentrations of NO₃ and N₂O₅ were 4.8 ± 3.3 pptv and 30.5 ± 35.4 pptv, respectively. The average NO₃ production rate was 1.03 ± 0.48 ppbvh⁻¹, and the steady-state lifetimes of NO₃ and N₂O₅ were 26 s and 162 s, respectively, indicating that the NOx chemistry in the rural site during summer was active. The uptake coefficient range of N₂O₅ was 0.023 to 0.118, with an average value of 0.062 ± 0.035. Meanwhile, the fitting for k_NO3 was 0.013 ± 0.016 s⁻¹, corresponding to the shorter NO₃ lifetime below 1 min. The results show that the indirect loss pathways caused by the heterogonous uptake of N₂O₅ contributed 64–90% of the overall NO₃ loss, with an average of 81%, suggesting that the N₂O₅ heterogeneous reaction dominated the nocturnal NO_x loss over this region.

1. Introduction

The nitrate radical (NO₃) and dinitrogen pentoxide (N₂O₅) are important reactive nitrogen species with regard to their nighttime chemistry. They play an important role in the nocturnal atmospheric chemistry, which controls the oxidation and removal of various trace gases [1]. O₃ reacts with NO₂ to form NO₃ (Equation (1)), and N₂O₅ is formed by the reaction of NO₃ with NO₂ (Equation (2)), which is in temperature-dependent equilibrium with NO₃ (Equation (3)) [2].

NO₂ + O₃ → NO₃ + O₂

(1)

NO₂ + NO₃+ M → N₂O₅ + M

(2)

N₂O₅+ M → NO₂ + NO₃+ M

(3)

NO₃ occurs during photolysis during the day and is accumulated only at night (up to several hundred pptv). The rapid reaction with NO is another important sink of NO₃ besides photolysis [3]. NO₃ can oxidize and remove certain volatile organic compounds (VOCs) and sulfides, and its oxidation ability even exceeds that of the daytime OH radical [4]. The terpenes released by natural plants react quickly with the NO₃ radical to form organic nitrate esters, which contributes to the formation of secondary organic aerosol (SOA) [5].

The heterogeneous hydrolysis of N₂O₅ forms soluble nitrate (Equation (4)), and heterogeneously reacts with Cl⁻ to form ClNO₂ and nitrate (NO₃⁻) (Equation (5)). The rate coefficient of the heterogeneous uptake of N₂O₅ is determined by Equation (6) [6]:

N₂O₅ + H₂O(l) → 2HNO₃

(4)

N₂O₅ + Cl⁻(aq) → ClNO₂ + NO₃⁻(aq)

(5)

k_N2O5 = 0.25 cS_aγ_N2O5

(6)

where k_N2O5 is the rate coefficient of the heterogeneous uptake of N₂O₅ and c is the mean molecular velocity of N₂O₅. Here, Sa represents the particle surface area and γ_N2O5 is the N₂O₅ uptake coefficient.

High mixing ratios of NO₃ and N₂O₅ have been observed in different environments, such as the Pearl River Delta [7], the North China Plain [8], and Hong Kong [9]. Studies have found that NO₃ and N₂O₅ have the characteristics of strong seasonality and regionality [10,11,12]. The initial studies on NO₃ mainly focused on its direct and indirect loss pathways [13,14]. Observations and model simulations have been used to study the heterogeneous N₂O₅ uptake, which is an important pathway of NO_X loss [15]. The N₂O₅ uptake coefficients in different environmental conditions were obtained to evaluate NO_X loss [16] and the contribution of nocturnal NO₃ to the formation of secondary organic aerosol has been studied [17]. In recent years, to investigate the nocturnal chemical processes of NOx, some field observations have been carried out in China. However, comprehensive field studies of N₂O₅ and NO₃ simultaneously have been relatively rare at rural sites on the North China Plain. On account of the importance of NO₃ and N₂O₅ in nocturnal atmospheric chemistry, it is of great interest to investigate the nocturnal chemical processes of NOx in the North China Plain.

In this study, self-developed cavity ring down spectroscopy (CRDS) was applied for the 2014 CAREBEIJING-NCP field experimental observation in summer (Wangdu, China). We report on observations of NO₃, N₂O₅, and relevant species. The factors influencing the production rate and lifetimes of NO₃ and N₂O₅ were deeply analyzed. In addition, heterogeneous N₂O₅ uptake coefficient was estimated and the loss mechanisms of NO₃ and N₂O₅ were further studied over this region during the summer time.

2. Method

2.1. Site Description

Cavity ring-down spectroscopy (CRDS) was applied for the Campaigns of Air Pollution Research in Megacity Beijing and North China Plain (CAREBEIJING-NCP) in Wangdu County, Hebei Province (38.665° N, 115.204° E) from 10 June to 6 July 2014, which was 170 km southwest of Beijing (Figure 1). The site was surrounded by crops, near the S032 and G4 expressways. The CRDS instrument was located on the fourth floor of the building, approximately 10 m above the ground (sample tubes of 1.1 m). As the measurement site was close to the national road, the NO₃ radical measurement results were susceptible to interference from vehicle emissions. During the observation period, northeasterly winds and easterly winds (from the direction of the G4 national highway) prevailed and the wind speed was relatively high, with a maximum wind speed of 26.5 m/s and an average wind speed of about 4.2 m/s (Figure 2).

2.2. Instruments

The principles of the CRDS instrument have been described previously [18,19]. The same CRDS instrument was applied in Wangdu, Hebei Province. The light from the laser diode with a center wavelength of 661.85 nm was coupled to an optical cavity. The cavity was formed using a pair of high-reflectivity mirrors (R ≥ 99.9985%). The light transmitted through the back mirrors was detected by a photomultiplier tube and then the signals were digitized using an oscilloscope card controlled by the LabVIEW program. The 1500 decay profiles were co-added to achieve a higher signal/noise ratio. Then, the decay trace was fitted by the Levenberg–Marquardt algorithm to obtain the ring-down time. The CRDS instrument was located approximately 10 m above the ground. The sample tubes of the CRDS instrument equaled 1.1 m and the sampling flow was 5 standard liters per minute. The detection limit of the NO₃ radical determined by Allan variance was approximately 1.6 pptv (1σ, 10 s). The uncertainty for NO₃ was about ±8%.

The relevant trace gases were measured simultaneously during the field observations. N₂O₅ was observed using a chemical ionization mass spectrometer (THS Instruments, Atlanta), and the detection limit was 7 pptv (3σ, 1 min). NO and NO₂ were obtained using a chemiluminescence analyzer (Thermo Fisher 42i, Waltham, MA, USA). O₃ was observed using an ultraviolet (UV) absorption analyzer (0.5 ppbv, 1 min). The meteorological parameters, including the wind direction, wind speed, relative humidity (RH), and temperature, were observed using an ultrasonic anemometer and a weather station, which were located 10 m above the ground. The aerosol surface (Sa) was obtained using a scanning mobility particle sizer (SMPS) and aerodynamic particle sizer (TSI-type APS model 3321).

3. Results and Discussion

3.1. Overview of Measurements

During the measurement period, the temperature and RH showed an obvious diurnal variation trend, and the variation trend was obviously opposite. When the atmospheric temperature was low at night, the RH reached the maximum. During the day, the temperatures peak around noon. During the observation period, the RH levels varied from 14% to 98%, and the average humidity was 62.7%. At the same time, the temperatures were higher, with an average temperature of 299.2 K, indicating that the whole campaign was conducted in high-temperature and -humidity environments.

The time series of NO₃ concentrations observed using the CRDS instrument and the relevant parameters is shown in Figure 3. During the observation period from June 10th to 6 July, the NO₃ nocturnal concentration varied from 0 to 25 pptv, and the average nocturnal concentration was about 4.8 ± 3.3 pptv. The highest NO₃ nocturnal concentration reached 25 pptv on 15 June and 27 June. There were prevailing northeasterly winds and easterly winds during the measurement period, and the measurement site was close to the national road and easily affected by local sources of vehicle emissions, which may have contributed to the relatively low NO₃ concentrations. The O₃ concentrations showed obvious periodicity, with peak concentration occurring around 16:00, with a maximum of 140 ppbv, which decreased to the detection limit after midnight. The O₃ concentrations were relatively high during the whole observation period, with an average concentration of 54 ppbv, which implied intense photochemical reactions during the observations. The NO₂ concentrations showed an inverse correlation with O₃, with an average value of 9.6 ppbv. During the measurement period, due to the influence of automobile emission sources, The NO concentrations were relatively high and several instantaneous peaks appeared during the observation period.

Here, the mixing ratio of N₂O₅ can be calculated using Equation (7). During the whole nocturnal observation period from June 10th to July 6th, the calculated N₂O₅ concentrations were relatively high, with a maximum of 374 pptv, which mainly attributed to the high mixing ratio of NO₃ and NO₂. The mixing ratio of observed N₂O₅ was measured via CIMS, which showed a high concentration during the period of 27 June to 29 June, with a maximum concentration of 429 pptv. The average concentration of observed N₂O₅ was 30.5 ± 35.4 pptv.

[N₂O₅] = Kep [NO₂][NO₃]

(7)

P(NO₃) = k_NO2+O3 [NO₂][O₃]

(8)

where k_NO2+O3 is the rate constant for the reaction of NO₂ with O₃ and Kep is the equilibrium constant dependent on the temperature.

3.2. NO₃ Production Rate and N₂O₅

Figure 4a shows the nocturnal profiles for the temperature, humidity, PM_2.5, PM₁₀, and Sa. During the nighttime period, the average temperature was 298 K and the maximum temperature appeared at 18:30. The nocturnal humidity increased gradually, with an average value of 68%. The PM_2.5 concentration was lower, below 50 μgm⁻³. Sa also presented a steadily rising trend, with an average value of 1687 μm²cm⁻³.

The NO₃ production rate (P(NO₃)) was calculated using Equation (8) [20]. The mean nocturnal variations of NO₂, O₃, N₂O₅, and P(NO₃) are shown in Figure 4b. During the nighttime period, the average concentrations of O₃ and NO₂ were 41.1 ± 24.1 ppbv and 10.9 ± 3.6 ppbv, respectively. The NO₂ and O₃ were highly inversely correlated. The O₃ concentrations decreased after sunset, while the NO₂ concentrations slightly increased. The P(NO₃) showed a similar trend with O₃, and was mainly affected by the O₃ variation. The peak of P(NO₃) appeared at 19:00–20:00 (an hour after the ozone peak), reaching 2.09 ppbvh⁻¹. The value for P(NO₃) was 1.03 ± 0.48 ppbvh⁻¹, which was slightly lower than that observed on a tower in Seoul (1.3 ppbvh⁻¹) [21] and in Beijing (1.2 ppbvh⁻¹) [22]. The average nighttime P(NO₃) value was slightly larger than that at a suburban site in Taizhou (1.01 ppbvh⁻¹) [17]. The comparison showed the significant oxidation capacity of NO₃ in Wangdu County.

The linear correlation coefficient between P(NO₃) and O₃ is plotted in Figure 5a. The average temperature was 297.7 K during the nighttime period, and the high temperature promoted NO₃ formation. The higher the temperature, the higher the P(NO₃), as shown in Figure 5a, which illustrated that the temperature was one of the main factors controlling the NO₃ formation. The other two factors, NO₂ and O₃, also determined the P(NO₃). The linear correlation coefficient between P(NO₃) and O₃ was 0.68, which was higher than for NO₂ (R = 0.06). The N₂O₅ concentration also showed a clear positive dependence on O₃ (Figure 5b). This result may imply that the P(NO₃) and N₂O₅ concentrations mainly varied with the O₃ concentrations, which could partly explain why O₃ varied more rapidly than NO₂ during the observation period.

During the measurement period, the mean nocturnal variations of the calculated N₂O₅ and observed N₂O₅ showed the same trend and their concentrations were basically consistent before midnight, with the large discrepancy shown after midnight. After midnight, the humidity was above 70 percent, which accelerated the heterogeneous hydrolysis of N₂O₅. The discrepancy between the calculated N₂O₅ and observed N₂O₅ was mainly ascribed to the heterogonous uptake of N₂O₅. The observed N₂O₅ concentration gradually increased after sunset, peaking near 22:30. After 00:00, P(NO₃) and N₂O₅ decreased simultaneously, suggesting that the low concentration of N₂O₅ was partly attributed to the lack of NO₃ formation.

3.3. Steady-State Lifetimes of NO₃ and N₂O₅

The steady-state lifetime was applied extensively when evaluating N₂O₅ and NO₃, and can be represented as in Equations (9) and (10) [23]:

$$\mathsf{\tau }\left({\mathrm{NO}}_3\right)=\frac{\left[{\mathrm{NO}}_3\right]}{{\mathrm{k}}_{{\mathrm{NO}}_2+{\mathrm{O}}_3}\left[{\mathrm{NO}}_2\right]\left[{\mathrm{O}}_3\right]}$$

(9)

$$\mathsf{\tau }\left({\mathrm{N}}_2{\mathrm{O}}_5\right)=\frac{\left[{\mathrm{N}}_2{\mathrm{O}}_5\right]}{{\mathrm{k}}_{{\mathrm{NO}}_2+{\mathrm{O}}_3}\left[{\mathrm{NO}}_2\right]\left[{\mathrm{O}}_3\right]}$$

(10)

During the whole campaign, the nocturnal N₂O₅ lifetime varied from 4 s to 13 min, with an average value of 162 s, while the average NO₃ lifetime was 26 s (Figure 6).

Table 1. The N₂O₅ concentration and P(NO₃) and $\mathsf{\tau }({\mathrm{N}}_{2 }{\mathrm{O}}_5$ ) values in Wangdu and other observation sites.

Location	Environment	N2O5 (pptv)	P(NO3) (ppbvh−1)	$\mathsf{\tau }({\mathbf{N}}_2{\mathbf{O}}_5)$	Reference
Beijing, China	Urban	<1100	1.2 ± 0.9	310 ± 240 s	[22]
Seoul, Korea	Urban, Tower	<5000	1.3	0.5 h	[21]
Hong Kong, China	Rural mountaintop	<8000	0.32	0.2–3 h	[9]
Taizhou, China	Suburban	<700	1.2 ± 0.4	55 ± 68 s	[24]
Wangdu, China	Rural	<500	1.03 ± 0.48	162 s	This study

shows the observed N₂O₅ concentration and P(NO₃) and $\mathsf{\tau }({\mathrm{N}}_2{\mathrm{O}}_5$) values in Wangdu and other observation sites. The lifetime of N₂O₅ in this study was slightly lower than the obtained N₂O₅ lifetime in Beijing (310 s, average) [22], but was much shorter than that in Hong Kong (0.2–3 h) [9] and that in Seoul, Korea (0.5 h, average) [21]. The shorter lifetime in Wangdu indicated that N₂O₅ and NO₃ were highly reactive and rapidly exhausted.

The influencing factors of the N₂O₅ lifetime were analyzed using the related parameters. When the RH exceeded 55%, the N₂O₅ lifetime showed a clear negative dependence on RH (Figure 7). The high RH effectively reduced the N₂O₅ lifetime, indicating that the heterogeneous N₂O₅ process played a large role as a sink. Figure 7b shows the correlation between the N₂O₅ lifetime and Sa. Here, the $\mathsf{\tau }({\mathrm{N}}_2{\mathrm{O}}_5$) rapidly decreased from 4 min to 1 min as the Sa increased up to 1250 μm² cm⁻³, but was approximately constant when the Sa increased from 1750 μm² cm⁻³ to 3750 μm² cm⁻³. The variation trend for the N₂O₅ lifetime with Sa indicated that the N₂O₅ uptake rates increased at higher Sa values. The N₂O₅ uptake was an important part of the N₂O₅ loss, which was promoted by the high RH and Sa.

3.4. NO₃ and N₂O₅ Loss Pathways

The loss of NO₃ was generally divided into two kinds, direct loss pathways caused by the reaction with the VOCs and NO and indirect loss pathways caused by N₂O₅ heterogeneous hydrolysis. To delve into the loss rates, the N₂O₅ uptake coefficient was estimated using steady-state methods (Equation (13)) [6]. The linear fits showed the N₂O₅ uptake coefficient, γ_N2O5, as the slope and the first-order loss rate coefficient for NO₃, k_NO3, as the intercept.

$$\mathsf{\tau }{(\mathrm{N}{\mathrm{O}}_3)}^{-1}={\mathrm{k}}_{{\mathrm{N}\mathrm{O}}_3}+{\mathrm{K}}_{\mathrm{eq}}\left[{\mathrm{NO}}_2\right]{\mathrm{k}}_{{\mathrm{N}}_2{\mathrm{O}}_5}$$

(11)

$$\mathsf{\tau }{({\mathrm{N}}_2{\mathrm{O}}_5)}^{-1}={\mathrm{k}}_{{\mathrm{N}}_2{\mathrm{O}}_5}+\frac{{\mathrm{k}}_{{\mathrm{NO}}_3}}{{\mathrm{K}}_{\mathrm{eq}}\left[{\mathrm{NO}}_2\right]}$$

(12)

$$\mathsf{\tau }{({\mathrm{N}}_2{\mathrm{O}}_5)}^{-1}{\mathrm{K}}_{\mathrm{eq}}\left[{\mathrm{NO}}_2\right]=\frac{1}{4}{\mathrm{cS}}_{\mathrm{a}}{\mathsf{\gamma }}_{{\mathrm{N}}_2{\mathrm{O}}_5}{\mathrm{K}}_{\mathrm{eq}}\left[{\mathrm{NO}}_2\right]+{\mathrm{k}}_{{\mathrm{NO}}_3}$$

(13)

In this study, which lasted only five nights, we obtained valid fitting results. The uptake coefficient range of N₂O₅ was 0.023 to 0.118, and the average uptake coefficient was 0.062 ± 0.035 (Figure 8). The high γ_N2O5 was associated with the RH and Sa. Compared with the other fields, the obtained N₂O₅ uptake coefficient in this study was larger than those observed in Hong Kong (0.014 ± 0.007) [9], New England (0.001–0.017) [20], Calgary (0.02) [25], and London (0.01–0.03) [26]. However, the value of γ_N2O5 was similar to that in Taizhou (0.027–0.107) [24].

The k_NO3 values varied from 0.0028 s⁻¹ to 0.041 s⁻¹, and the average was 0.013 ± 0.016 s⁻¹, corresponding to the shorter NO₃ lifetime below 1 min (

Table 2. The N₂O₅ uptake coefficient_, k(_NO3), and k_N2O5.

Date	Sa (μm2/cm3)	γN2O5	KN2O5 (s−1)	KNO3 (s−1)
20–21 June	1876	0.068	0.0077	0.041
27–28 June	1185	0.023	0.0016	0.013
27–28 June	926	0.051	0.0027	0.0028
29–30 June	1090	0.052	0.0034	0.0041
30 June–1 July	1077	0.118	0.0077	0.0055

). The high k_NO3 indicated that the reactions of NO₃ with NO and the VOCs represented important sinks. The NO₃ removal process caused by the reactions with the VOCs and NO accounted for about 10–36% of the total during the five nighttime cases. The rate coefficient of the heterogeneous uptake of N₂O₅ (k_N2O5) can be calculated using Equation (6). The values of k_N2O5 varied from 0.0016 s⁻¹ to 0.077 s⁻¹, with an average value of 0.046 ± 0.029 s⁻¹ during the five nighttime cases. The indirect loss pathways caused by the heterogonous uptake of N₂O₅ were estimated, which contributed 64–90% of the overall NO₃ loss (average of 81%), suggesting the dominance of the heterogeneous reaction of N₂O₅ in nocturnal NO₃ loss. Compared with the studies in South China, the loss of NOx was dominated by the VOCs followed by N₂O₅ uptake in Tai’Zhou in polluted Southern China [17].

4. Conclusions

During the measurement period, the maximum concentrations of nocturnal NO₃ and N₂O₅ were 25 pptv and 429 pptv. The P(NO₃) varied mainly with the ambient O₃ concentrations. The observed N₂O₅ concentration gradually increased after sunset, mostly peaking near 22:30. The nocturnal average lifetimes of NO₃ and N₂O₅ were 26 s and 162 s, which were 3-fold smaller than that reported in Beijing, The shorter lifetime in Wangdu indicated that the N₂O₅ and NO₃ were highly reactive and rapidly exhausted over this region in the North China Plain. The estimated heterogeneous N₂O₅ uptake coefficient range was 0.023 to 0.118 during the five nighttime cases. The indirect loss by the heterogonous N₂O₅ uptake was estimated, which contributed 64–90% of the overall NO₃ loss. The N₂O₅ lifetime had a clear negative dependence on the RH and Sa. The high RH and high Sa may have promoted the heterogonous N₂O₅ uptake during the campaign.

Without the quantification of the contribution of the VOCs to NOx loss during the campaign, the above analysis based on the N₂O₅ uptake coefficient may have some uncertainties. The results in this work demonstrate the importance of the heterogeneous chemistry in nocturnal NO₃ and N₂O₅ removal in rural sites on the North China Plain. Meanwhile, it may be applicable to the N₂O₅ uptake processes in similar regions with high NOx, high RH, and high Sa.