Project Topics
Get Complete Project Material File(s) Now! »
Development of a stripping coil-ion chromatograph method to measure atmospheric HONO
To summarize, this chapter described a stripping coil (SC) equipped with ion chromatograph (IC) to measure HONO, which was developed and assessed. Briefly, the reliability of the method mainly depends on the collection efficiency and the interference with other species. The performance of the method was assessed in the chamber using two kinds of absorption solutions, i.e., ultrapure water and 25 M Na2CO3 solution under different concentrations of SO2. Results indicated that HONO concentrations absorbed by ultrapure water and Na2CO3 solution were almost identical in the absence of SO2 in the chamber, and both the collection efficiencies were more than 99%. However, the collection efficiency of ultrapure water decreased with the increase of SO2, indicating that the presence of SO2 resulted in the penetration of HONO during the sampling process in water. The collection efficiency kept more than 90% when the concentration of SO2 was no more than 23 ppbv. Comparing with the situation without SO2, HONO performed a remarkable increase with the presence of SO2 when using Na2CO3 absorption solution, indicating the extra generation of HONO from the reaction between SO2 and NO2 in alkaline solution. Consequently, using ultrapure water as the absorption solution could provide a high collection efficiency and avoid the interferences from SO2 when the concentration of SO2 was below 23 ppbv. High correlations (slope=0.94~1.06, r2>0.90) were found during the intercomparisons between SC-IC and other three techniques (two LOPAPs and one CEAS), suggesting the SC-IC method developed in this study was able to measure atmospheric HONO in the field campaigns.
Experimental
Technical setup of the SC-IC
In brief, the SC-IC system was based on the wet chemical method. Atmospheric HONO was absorbed by the absorption solution (ultrapure water or Na2CO3 solution) and converted to NO2- in the liquid samples. NO2- in the liquid samples was quantified by IC. Then the atmospheric HONO mixing ratio could be calculated by the gas sampling flow, liquid flow, NO2- concentrations in the liquid samples, and atmospheric temperature.
As shown in Figure 2.1, the SC-IC device consists of three units: the sampling unit (part A), the transferring and supporting unit (part B), and the detection unit (part C). The sampling unit consists of a five-turn stripping coil (2 mm inner diameter), which is enclosed by a glass cylinder. The stripping coil has two inlets in the front (one for air and another is for absorption solution) and a gas-liquid separator. A small non-transparent box was used to cover the stripping coil to protect the SC and keep it away from sunlight with the air inlet stretching out the box for about 2 cm. A circulating water bath is used to keep the temperature of the stripping coil stable at 20 1 C. Consequently, the sampling unit is conveniently movable. The sampling unit can be placed outside without inlet tubes in the field measurements, avoiding interference from the heterogeneous reaction on the surface of the inlet tubes.
The transferring and supporting unit consists of a dryer, a circulating water bath, a mass flow controller (MFC), an air membrane pump (KNF, Germany), two peristaltic pumps (Shenchen, China), a micro-filter, a glass bottle (1 L) for absorption solution, a 24-port valve and some sample bottles (20 mL). During the sampling process, the ambient air is drawn into the stripping coil by the air membrane pump. The air is absolutely mixed with the absorption solution supplied by the peristaltic pump and then the solution in the stripping coil is gathered in the gas-liquid separator. In the gas-liquid separator, the air is continually drawn through a dryer and an MFC and finally released as exhaust gas by the air membrane pump. The liquid is pumped through the micro-filter and the 24-port valve and collected in the sample bottles by another peristaltic pump. The MFC is weekly calibrated by a soap-foam flowmeter.
As the liquid volume for IC to analyze should be no less than 0.5 mL, the method we designed is able to meet the high time resolution (>2.5 min for one sample) for field measurements or laboratory researches.
Collection efficiency
where H is the Henry constant of HONO (49 M atm-1), Ka represents the ionization constant of HONO, and [H+] is the acidity of the absorption solution. Consequently, the collection efficiency mainly depends on the ratio α= Fg/Fl and the acidity. A proper α is needed because too large Fg/Fl results in a low β and too small α one results in more residence time, which enlarges the potential interference from the heterogeneous reactions. Also, proper acidity is needed because high acidity (low pH) leads to a low β and low acidity (high pH) is in favor of the interference produced from heterogenous or multiphase reactions of NO2.
Besides the ideal collection efficiency β, the measured βm, obtained from two stripping coils connected in series, is defined in Eq (2.4).
Laboratory tests and field measurements
To explore the collection efficiency and potential interference of the sample system, laboratory researches have been conducted in a chamber. The chamber is made of Teflon film with a volume of 3.6 m3. The temperature of the chamber is controlled to be 25 C by an air conditioner. The chamber is inside a stainless box for light shielding to control the radiation in the chamber.
N2 was introduced into the chamber first, and then standard gas of NO2 was injected into the chamber to generate NO2 concentration of 100 ppbv. Then, gradient concentrations of SO2 (0-113ppbv) were achieved by injecting SO2 standard gas orderly. Two stripping coils connected in series are used to collect HONO in the chamber. The impact on HONO collection efficiency from SO2 can be reflected by the change of β.
To investigate the impact on HONO collection efficiency from different absorption solutions, the researches just mentioned before are conducted by two sample systems: one uses the ultrapure water as the absorption solution (H-method) and the other uses the 25 M Na2CO3 solution (N-method). The gas flow for sampling and the liquid flow for absorption solution are set as 2 L min-1 and 0.2 mL min-1, respectively. Four replicate samples for each sample system are collected. Each sample period lasts 5 min. To explore the impact from particulate nitrite, atmospheric particles are collected on the quartz membrane by a particle sampler (Laoying, China) in Beijing from 3 January 2016 to 21 January 2016 during the winter field measurement. The sampling site is on the rooftop of a building in Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences. The site is well documented by previous studies128. During the measurement, particles were collected every 2 hours at a sample flow of 100 L min-1.
To test the performance of SC-IC on measuring atmospheric HONO in the field measurement, comparison researches were conducted with other methods including CEAS and two LOPAPs at the Station of Rural Environment, Chinese Academy of Sciences (SRE-CAS) located in DongBaiTuo village (38 71 N, 115 15 E), Hebei Province of China. LOPAP-1 was homemade by the Institute of Chemistry, Chinese Academy of Sciences. It has been proved to work well in many field measurements. The comparison with LOPAP-1 lasted from 13 June 2017 to 20 June 2017. LOPAP-2 was a commercial product of QUMA Elektronik & Analytik GmbH, Germany. The comparison with LOPAP-2 was conducted from 10 November 2017 to 22 November 2017. The CEAS was homemade by Anhui Institute of Optics and Fine Mechanics, Chinese Academy of Sciences. The comparison with CEAS lasted from 15 June 2017 to 22 June 2017. Note that both the sample inlets of SC-IC and CEAS were in a dynamic chamber which was serviced for measuring HONO emission flux from fertilized agricultural soil.
During the three comparisons, the SC-IC and LOPAPs ware calibrated by the same standard NaNO2 solution. The CEAS was calibrated once per day. A 24-port valve (Figure 2.1) was used in SC-IC to achieve automatic collection, and the switching frequency was set at once per hour. Therefore, the hourly average of data from CEAS and LOPAPs were used for comparisons.
Results and discussion
Impact from Fg/Fl
Theoretically, the ideal collection efficiency decreases with the increase of Fg/Fl (α) (Eq (2.2)). However, too small α results in long residence time, enlarging the influence of heterogeneous reaction. Hence, it needs to conduct field measurements to choose proper α. So, two stripping coils were connected in series to measure atmospheric HONO and ultrapure water was used as the absorption solution in a heavy haze winter day of Beijing. Six different sample flows (0.5-3 L min-1) were set with Fl at 0.2 mL min-1, corresponding to α of 2500-12500. Considering the solubility equilibrium of atmospheric CO2 in water, the pH of ultrapure water is 5.6. Then some studies used the 25 M Na2CO3 solution to balance the equilibrium of atmospheric CO2 (pH=6.9). Besides the measured collection efficiency βm, the ideal collection efficiency at pH=5.6 (βpH=5.6) and pH=6.9 (βpH=6.9) was also calculated (Figure 2.2). It follows that HONO concentration in the second stripping coil under the sample flow above 2 L min-1 (2.5 and 3 L min-1) was obviously higher than that under the sample flow below 2 L min-1, indicating that large airflow or Fg/Fl (>12500) would result in high penetration proportion. While the concentrations of atmospheric SO2 in each sample flow period were 12 (3 L min-1), 11 (2.5 L min-1), 14 (2 L min-1), 11 (1.5 L min-1), 14 (1 L min-1) and 14 ppbv (0.5 L min-1), respectively, the large penetration rate was not ascribed to the increase of SO2 but the limited absorption time.
Although the collection efficiency was expected to be higher at low α values, βm was remarkably lower (<92%) than the ideal collection efficiency at the Fl of 0.5 or 1.0 L min-1 (α=2500 or 5000). On the one hand, a small gas flow might introduce more heterogeneous reactions, as discussed before. It was obvious that the absorption solution in the stripping coil distributed quite inhomogeneously either at a large Fg (>2 L min-1) or low Fg (<1.0 L min-1), the effective collision between gas molecules and liquid surface reduced sharply, resulting in a low βm. Besides, the disturbance in the inhomogeneous situation caused the distinctly significant fluctuation (error bars) of the collection efficiency (Figure 2.2).
Compared with the situation under other sample flows, the sampling system showed better performance with the sample flows of 1.5 and 2 L min-1. It can not only collect HONO at a high collection efficiency of more than 94.1% but also possess a short residence time (about 0.1 s) to avoid the heterogeneous contribution. Therefore, the sample flow of 1.5~2 L min-1, corresponding to the Fg/Fl of 7500-10000, was recommended to accurately measure atmospheric HONO for the SC-IC method.
Impact from NO2 and SO2
Two stripping coils connected in series were used in the chamber experiments to assess the collection efficiency. The collection efficiency and the concentrations of HONO in the two stripping coils were measured when the chamber was filled with N2 and 100 ppbv NO2 with different concentrations of SO2. HONO concentration in the chamber filled with N2 was less than 5 pptv (Figure 2.3A). However, when NO2 standard gas was injected into the chamber to get its concentration up to 100 ppbv in the chamber, averaged HONO concentration of 20 samples reached up to 2.235 0.027 ppbv, indicating that the injection of NO2 increased HONO in the chamber. Due to the small Henry constant (9.87 10-8 mol L-1 Pa at 25 C in water) of NO2, the concentrations of NO2 in the first and second stripping coils were almost the same. However, HONO in the first stripping coil was two orders of magnitude larger than that in the second, suggesting that the contribution from NO2 reactions on the surface, including heterogeneous reactions to the measured HONO was negligible. Therefore, the increase of HONO after injecting NO2 was ascribed to two potential reasons: 1) HONO existed in the standard gas, and 2) HONO was produced from the NO2 heterogeneous reactions on the chamber wall. Cheng et al. (2013)128 used dynamic dilution to achieve 200 ppbv NO2 and found only 4 pptv HONO in the airflow measured by the SC-IC method, indicating that HONO came from the heterogeneous reaction of NO2 on the chamber wall. If the reaction worked continuously, HONO would increase with the reaction time. Whereas, HONO concentration kept stable during a long period (>120min) after injecting NO2, suggesting a large initial uptake coefficient and a small uptake coefficient when the gas was well mixed. In addition, HONO concentrations measured by H-method or N-method were almost identical after injecting NO2, and the penetration rate was less than 1% (Figure 2.3B), indicating that both methods were able to collect HONO with high collection efficiency in this condition.
Compared with the situation when only NO2 was injected into the chamber, HONO concentrations in the first stripping coil increased remarkably using N-method but decreased gradually using H-method after SO2 was injected into the chamber (Figure 2.3A). Previous researches have demonstrated that NO2 could oxidize SO2 on the liquid surface through heterogeneous reaction or multiphase reaction with the production of HONO (Eq (2.5)). The reaction was in favor of alkaline conditions such as the high ammonia areas.
Obviously, in the situation where certain concentrations of NO2 and SO2 co-existed, HONO concentration would be overestimated because of the HONO production from Eq (2.5), especially for N-method. Even if the penetration percent of H-method increased with the concentration of SO2, it was no more than 8% when SO2 was below 23 ppbv. When SO2 in the chamber reached up to 113 ppbv, the concentration of HONO in the first stripping coil in N-method was consistent with the initial HONO detected just after injecting NO2, which would be ascribed to the decrease of HONO production and the increase of penetration percent. SO2, as a kind of acid gas, can neutralize the alkalinity of the absorption solution. On the one hand, high concentration of SO2 reduced the heterogeneous production of HONO; on the other hand, it reduced the solubility of HONO, resulting in a high penetration percent (Figure 2.3). Therefore, it was a coincidence that HONO in the first stripping coil in N-method was consistent with the initial HONO concentration. Cheng et al. have also explored the impact of SO2 on HONO measurement using the SC-IC method, and they found the impact can be neglected. Note that their result only based on the situation when 100 ppbv SO2 and 300 ppbv NO2 were co-existed. However, our research found that the heterogeneous production of HONO can be just counteracted by the penetration with SO2 concentration of 113 ppbv and NO2 concentration of 100 ppbv.
In the typical polluted areas, including Beijing, the concentration of SO2 is usually about several ppbv in summer and tens of ppbv in winter. Based on our research, H-method has the capacity to carry on HONO field measurements in the typical polluted areas of China. The method can not only provide a high collection efficiency but also reduce the interference from the co-existence of SO2 and NO2. In addition, the study recommended that systematic assessments about the absorption efficiency and potential interference should be conducted when using wet chemical methods to detect HONO.
Particulate nitrite and storage time
As our samples are collected by offline method, they will be stored for some time (usually less than 10 days) before analysis. We explored the impact of storage time on HONO measurement. We collected HONO using H-method and conducted the analysis immediately. Liquid samples were kept hermetic and stored in a refrigerator at 4 C for 11 days (264 hours). After that, HONO in the samples was analyzed again to compare with that before storage with the results shown in Figure 2.4A. It’s evident that HONO concentrations before and after storage show a high correlation (slope=1.005, r2=0.997), suggesting that the liquid sample possesses enough chemical stability for at least 11 days when kept at 4 C.
In addition, to explore the stability of the sample in the environmental temperature, we picked 12 samples with different HONO concentrations from the campaign in winter 2015/2016. We analyzed them every 2 hours in 24 hours (Figure 2.4B). Results showed that HONO performed no significant variation in all the samples. Statistical analysis revealed that the relative standard deviation was less than 14.3% when HONO concentration was below 0.2 ppbv and less than 1.7% when HONO concentration was more than 0.2 ppbv. The variation might be caused by the uncertainty of detecting NO2-by the ion chromatograph rather than the storage time.
Intercomparisons
To further evaluate the reliability of the SC-IC developed in this study, field comparisons between the SC-IC and the methods of LOPAP and CEAS were conducted. As shown in Figure 2.5, the levels and the variations trends of HONO measured by the SC-IC were in good agreement with those measured by the LOPAP-1 (Figure 2.5a), LOPAP-2 (Figure 2.5b) and CEAS (Figure 2.5c) during the three measurement periods, which could be reflected by the significant correlations (slopes 1 and r2 >0.90) between the HONO concentrations measured by SC-IC and the other two methods (Figure 2.5d). The approximately same parameters (Mean, Median, Standard Deviation, Minimum, Maximum) from the summary statistics (Table 2.1) for each comparison revealed that the SC-IC method developed in this study was able to measure atmospheric HONO from low concentration (0.1 ppbv) to high concentration (14 ppbv).
Table of contents :
Chapter 1 Introduction: atmospheric chemistry of HONO
1.1 The role of HONO in atmospheric chemistry
1.1.1 Hydroxyl radicals (OH)
1.1.2 The contribution of HONO to OH
1.2 The proposed HONO sources and sinks
1.2.1 Direct emissions
1.2.2 Homogeneous reactions
1.2.3 Heterogeneous conversion of NO2
1.2.4 Photolysis of nitric acid and particulate nitrate
1.2.5 Acid displacement of soil surface nitrite by strong atmospheric acids
1.2.6 Soil HONO emissions
1.2.7 Other HONO sources
1.3 HONO measurement techniques
1.4 HONO measurements in the NCP region
Chapter 2 Development of a stripping coil-ion chromatograph method to measure atmospheric HONO
2.1 Experimental
2.1.1 Technical setup of the SC-IC
2.1.2 Collection efficiency
2.1.3 Laboratory tests and field measurements
2.2 Results and discussion
2.2.1 Impact from Fg/Fl
2.2.2 Impact from NO2 and SO2
2.2.3 Particulate nitrite and storage time
2.2.4 Intercomparisons
2.3 Summary
Chapter 3 Development of a twin open-top chambers method to measure soil HONO emission flux
3.1 Experimental
3.1.1 Twin open-top chambers method (OTCs)
3.1.2 HONO collection and analyzer
3.1.3 Laboratory research about the performance of OTCs
3.1.4 Site description
3.2 Results and discussion
3.2.1 Performance of the OTCs system in the laboratory research
3.2.2 Performance of the OTCs system in the field test
3.2.3 Greenhouse effect in the chamber
3.2.4 Influence from chemical reactions in the chamber
3.2.5 HONO emissions before fertilization
3.2.6 HONO emissions after fertilization
3.3 Summary
Chapter 4 Soil HONO emission flux measurement and regional O3 pollution in the summertime
4.1 Field evidence for soil HONO emissions
4.1.1 Atmospheric HONO and related parameters measurement
4.1.2 Photolysis frequency values (J)
4.1.3 Overview of the observations
4.1.4 Discussion on soil HONO emissions and their impact
4.2 Soil HONO emission flux measurement
4.2.1 Experimental
4.2.2 Overview of the measurements
4.2.3 Possible mechanisms
4.3 Laboratory experiments on soil HONO emission mechanism
4.3.1 Soil samples
4.3.2 Flow tube
4.3.3 Treatments of the soil samples
4.3.4 Results of a new mechanism
4.4 Regional impact on O3 pollution
4.4.1 CMAQ model configurations
4.4.2 Simulation cases
4.4.3 Impacts of soil HONO emissions on HONO and O3 concentrations at the Wangdu site
4.4.4 Regional impacts of soil HONO emissions in the NCP region
4.5 Reduction strategy
4.6 Summary and conclusions
Chapter 5 HONO budget and its role in nitrate formation in the wintertime
5.1 Field measurements
5.1.1 Site description and instrumentation
5.1.2 NO2 correction
5.2 Model description
5.2.1 MCM .
5.2.2 Model Configurations
5.2.3 Parameterization of HONO sources/sinks
5.3 Results and discussion
5.3.1 Overview of the observations
5.3.2 OH simulations
5.3.3 HONO simulations and budget
5.3.4 Implications on HOx chemistry and nitrate formation
5.4 Summary
