ATMOSPHERIC AEROSOL INORGANIC AND ORGANIC NITROGEN QUANTIFICATION METHOD AND SYSTEM

Abstract

A method of atmospheric inorganic and organic nitrogen quantification is disclosed. The ambient air is sampled by drawing it through an inlet followed by a denuder to reduce positive artifacts. After artifact removal, the air sample is collected onto a filter. The filter is subjected to thermal evolution under stepwise temperature program to generate a gaseous product mixture. In the presence of oxygen-containing carrier gas, the gaseous product mixture is oxidized to form oxidized gaseous products of CO.sub.2 and nitrogen oxides. Then, the nitrogen oxides products are processed to form an NO product and reacted with ozone to form an excited NO.sub.2* molecule. By quantifying the intensity of fluorescence, the concentration of NO.sub.2* molecule is measured, which determines the nitrogen content in the aerosol sample. The differentiation of inorganic and organic nitrogen is achieved through processing the thermally evolved carbon and nitrogen signals using multivariate curve resolution data treatment.

Claims

1. An atmospheric aerosol inorganic and organic nitrogen quantification method, comprising: sampling an ambient air by drawing it though an inlet to obtain an air sample; reducing positive artifacts of the air sample; filtering the air sample to collect an aerosol product; thermally evolving the aerosol product through a stepwise temperature program to obtain a gaseous product mixture; oxidizing the gaseous product mixture with an oxygen-containing carrier gas to generate oxidized gaseous products; forming an NO product from the oxidized gaseous products; reacting the NO product with ozone to produce an excited NO.sub.2* molecule for causing a chemiluminescent reaction; and detecting and quantifying a fluorescence emitted during the chemiluminescent reaction to obtain an intensity value of the fluorescence, wherein the intensity value of the fluorescence represents a total nitrogen mass concentration as a N signal of the ambient nitrogenous aerosols. processing the evolved carbon (C) and nitrogen (N) signals using a multivariate curve resolution (MCR) tool to resolve overlapping peaks.

2. The method of claim 1, wherein the inlet has a cut cyclone and the ambient air is drawn at a flow rate of at least 8.0 L/min.

3. The method of claim 1, wherein the positive artifacts are caused by adsorption of gas-phase compounds and reduced by a parallel plate filter denuder.

4. The method of claim 1, wherein the stepwise temperature program is a 6-step temperature program for thermal evolution.

5. The method of claim 1, wherein the 6-step temperature program is a heating program with 6 progressive heating temperature degrees, including approximately 150° C., 180° C., 300° C., 400° C., 500° C. and 800° C.

6. The method of claim 1, wherein the gaseous product mixture is oxidized to the oxidized gaseous products under 840-870° C. in the presence of oxygen and MnO.sub.2 catalyst.

7. The method of claim 6, wherein the oxidized gaseous products comprise a CO.sub.2 product and nitrogen oxides products.

8. The method of claim 7, wherein the oxidized gaseous products are split and subjected into two paths, wherein one path is directed to a carbon detector for detecting a carbon amount and generating a C signal and the other path is directed to a NO.sub.x analyzer comprising a NO convertor for transforming the nitrogen oxides products into the NO product.

9. The method of claim 8, wherein the carbon detector is a non-dispersive infra-red (NDIR) detector that monitors CO.sub.2 product and measure the carbon amount of the CO.sub.2 product to generate the C signal or a flame ionization detector (FID) that converts the CO.sub.2 product to a CH.sub.4 product and measure the carbon amount of the CH.sub.4 product to generate the C signal.

10. The method of claim 9, wherein a C/N signal ratio is calculated to evaluate whether the N signal is dominated by inorganic nitrogen or organic nitrogen.

11. The method of claim 10, wherein the C signal and N signal are processed resolve overlapping peaks by using a multivariate curve resolution (MCR) tool.

12. The method of claim 8, wherein the NO convertor is a molybdenum convertor.

13. The method of claim 1, wherein a flow-adjustable needle valve is deployed on the path to the NO.sub.x analyzer.

14. The method of claim 1, wherein a calibration curve is constructed by a standard solution containing both carbon and nitrogen.

15. An atmospheric aerosol inorganic and organic nitrogen quantification system, comprising: an inlet for intaking atmospheric air samples; a denuder for reducing positive artifacts to obtain an aerosol product; a carbon analyzer having a front oven for thermally evolving the aerosol product to obtain a gaseous product mixture and a back oven for oxidizing the gaseous product mixture with an oxygen-containing carrier gas to generate oxidized gaseous products; a carbon detector for carbon measurement; a needle valve for flow adjustment; a NO.sub.x analyzer having a NO convertor; an ozonator for generating ozone; a reaction chamber for reacting an NO product with ozone; a fluorescence quantification module for evaluating and quantifying fluorescence intensity and correlating the fluorescence intensity to a concentration of NO.sub.2* molecules.

16. The system of claim 15, wherein the inlet has a cut cyclone and the ambient air is drawn at a flow rate of at least 8.0 L/min.

17. The system of claim 15, wherein the denuder is a parallel plate filter denuder.

18. The system of claim 15, wherein the front oven performs a stepwise temperature program.

19. The system of claim 15, wherein the oxygen-containing carrier gas is He/O.sub.2.

20. The system of claim 15, wherein the catalytic convertor is a molybdenum catalytic convertor.

21. The system of claim 15, wherein the carbon detector is a nondispersive infrared detector or a flame ionization detector.

22. The system of claim 15, wherein the system further comprises a data logging module for analyzing, calculating, and saving the fluorescence intensity data, a temperature control module for adjusting and controlling the stepwise temperature program of the front oven and the temperature of back oven, and a controller to control the data logging module and the temperature control module.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Embodiments of the invention are described in more detail hereinafter with reference to the drawings, in which:

[0033] FIG. 1 depicts nitrogen recovery test results of different inorganic and organic nitrogen standards;

[0034] FIGS. 2A-2E depict typical nitrogen thermograms under the 6-step temperature program and the corresponding nitrogen masses of nitrogen standards and ambient samples; FIG. 2A depicts the thermograms of NH.sub.4NO.sub.3, (NH.sub.4).sub.2SO.sub.4 and NH.sub.4HSO.sub.4; FIG. 2B shows the nitrogen mass distribution of NH.sub.4NO.sub.3, (NH.sub.4).sub.2SO.sub.4 and NH.sub.4HSO.sub.4; FIG. 2C shows the thermogram and the stepwise nitrogen mass distributions of HULIS extracted from ambient PM.sub.2.5 sample, representing the thermal evolution characteristics of water-soluble ON complex mixture; FIG. 2D depicts the thermogram of a whole ambient aerosol sample; and FIG. 2E shows the stepwise nitrogen mass distribution of a whole ambient aerosol sample;

[0035] FIGS. 3A-3K shows example C and N thermograms under a 6-step temperature program and corresponding distributions of N masses at each temperature step; FIG. 3A depicts 6-step temperature program (150, 180, 300, 400, 500 and 800° C., where the first two temperatures are repeated to circumvent the instrument limit on the longest residence time allowed (i.e., 400 s) while meeting the need to have a heated period longer than 400 s and signals from the repeat steps of the same temperature are combined in data analysis; FIG. 3B is the thermogram of a single standard of NH.sub.4NO3; FIG. 3C shows the N mass distribution of NH.sub.4NO.sub.3 at each temperature step; FIG. 3D exhibits the thermograms of single standards of (NH.sub.4).sub.2SO.sub.4 and NH.sub.4HSO.sub.4; FIG. 3E shows the N mass distributions of (NH.sub.4).sub.2SO.sub.4 and NH.sub.4HSO.sub.4 at each temperature step; FIG. 3F is the thermogram of an aerosol WSON complex mixture (HULIS); FIG. 3G shows the N mass distribution of a HULIS sample at each temperature step; FIG. 3H is the thermogram of an aerosol WION complex mixture; FIG. 3I depicts the N mass distribution of a WION sample at each temperature step. FIG. 3J is the thermogram of a whole ambient aerosol sample; and FIG. 3K shows the N mass distribution of an ambient sample at each temperature step;

[0036] FIG. 4 depicts a schematic diagram of an online atmospheric inorganic and organic nitrogen quantification system;

[0037] FIG. 5 shows the conceptual illustration of unraveling the mixed thermally evolved IN and ON signals to pure IN and ON components by a multivariate curve resolution tool; m is a matrix of sample data with mixed IN and ON signals, C and P represent the factor contribution matrix and factor profile matrix, respectively, showing the information of resolved IN and ON components, a, b, and c indicate the size of the matrixes;

[0038] FIG. 6 shows time series of NO.sub.x, PM.sub.2.5, TN, IN, and ON concentrations at the Hong Kong University of Science and Technology (HKUST) site during the measurement period (29 Nov. 2020-1 May 2021);

[0039] FIGS. 7A-7C depict time series of meteorological conditions, gaseous pollutants, organic molecular tracer species and aerosol inorganic nitrogen and organic nitrogen concentrations at the HKUST site during three case periods, respectively; Levo means levoglucosan, α-PinT stands for α-pinene secondary organic aerosol (SOA) tracers, IsoT is isoprene SOA tracers, β-CaryT means β-caryophyllene SOA tracers; 4-NC is 4-nitrocatechol; 4-NP stands for 4-nitrophenol;

[0040] FIGS. 8A-8D are scatter plots of aerosol ON versus different nitrogen-containing species; FIG. 8A depicts aerosol ON versus NO.sub.x; FIG. 8B depicts aerosol ON versus IN; FIG. 8C exhibits aerosol ON versus NH.sub.4.sup.+—N; and FIG. 8D shows aerosol ON versus NO.sub.3.sup.−—N; and

[0041] FIGS. 9A-9C show comparisons between the aerosol IN concentrations measured by the present online inorganic and organic nitrogen quantification method and the traditional ion chromatography (IC) analysis; FIG. 9A depicts the comparison between the present online method and the offline IC analysis of quartz filter; FIG. 9B depicts the comparison between the present online method and the offline IC analysis of Teflon filter; and FIG. 9C shows the correlations between the offline IC analysis of quartz filter and Teflon filter.

DETAILED DESCRIPTION

[0042] In the following description, a method of both offline laboratory and online atmospheric inorganic and organic nitrogen quantification and a system thereof are set forth as preferred examples. It will be apparent to those skilled in the art that modifications, including additions and/or substitutions may be made without departing from the scope and spirit of the invention. Specific details may be omitted so as not to obscure the invention; however, the disclosure is written to enable one skilled in the art to practice the teachings herein without undue experimentation.

[0043] In accordance with a first aspect of the present invention, the present invention provides a method of atmospheric bulk inorganic and organic nitrogen quantification. The online atmospheric nitrogen quantification method integrates an online aerosol carbon analyzer based on thermal evolution method (RT-3131, Sunset Laboratory, USA) and a chemiluminescent NO.sub.x analyzer (Model 42i, Thermo Scientific, USA). For evaluating aerosol inorganic and organic nitrogen in atmosphere, the ambient air is sampled. During the sampling period, the ambient air is drawn through an inlet with a 2.5 μm cut cyclone at a flow rate of 8.0 L/min. A parallel-plate filter denuder (Sunset Laboratory, USA) is installed to reduce the positive artifacts from the gas-phase compound adsorption to the quartz fiber filter. In the front oven of the carbon analyzer, ambient PM.sub.2.5 is collected on a 1.23 cm.sup.2 circular quartz filter pre-baked during the previous analysis cycle. During the analysis, aerosols collected on the filter are thermally evolved under the stepwise temperature program, and then enter the back oven with the He/O.sub.2 carrier gas. In the back oven, the gaseous products are oxidized to CO.sub.2 and nitrogen oxides (NO.sub.y) under 840° C. in the presence of oxygen and MnO.sub.2 catalyst. Then, the flow is split into two paths, one to an NDIR detector to monitor C signal and the other to a NO.sub.x analyzer. The NO.sub.y products enter the NO.sub.x analyzer and are transformed into NO by the molybdenum (Mo) catalytic convertor. In the reaction chamber, the ozonator generates ozone to react with NO and produces excited NO.sub.2* molecules. The fluorescence light emitted during this chemiluminescent reaction is detected by a photomultiplier tube for quantification of nitrogenous species. Calibration curve is constructed by spiking standard solutions containing C and N onto the prebaked filter. Thus, total nitrogen mass concentrations of the analyzed aerosol sample can be quantified directly.

[0044] The evolved carbon (C) and nitrogen (N) products are monitored simultaneously. The joint monitoring of the CO.sub.2 product and nitrogen oxides products is an important feature of the analyzer. This feature facilitates differentiate IN and ON as IN species only have N signal while ON materials produce both C and N signals. The N signal as a function of the temperature program produces a graph termed as a N thermogram. Similarly, the C signal as a function of the temperature program produces a graph termed as a C thermogram. In the aerosol N thermograms, while IN and ON signals exhibit distinct thermal characteristics, they are not distinctive peaks. A multivariate curve resolution (MCR) tool is used to solve the overlapping peaks. The MCR data treatment is implemented using Positive Matrix Factorization (PMF) model. Specifically, PMF is performed on a data matrix X (n×m), in which n is the number of samples, m is the number of C and N peaks in the thermograms, and a given entry Xi in the matrix is the peak intensity of C or N of a peak in the thermogram. The PMF model decomposes the data matrix X into two matrices: factor contributions and factor profiles. As a result, the IN and ON factors (i.e., subcomponents of IN and ON) are resolved relying on their distinct thermal profiles. The IN and ON quantities are then obtained by summing up the N masses in all IN and ON factors, respectively.

[0045] In another embodiment of the present invention, it can be used as a laboratory equipment for evaluating the sample taken from several places. For such, an offline atmospheric aerosol inorganic and organic nitrogen quantification system is provided. The offline system shares the same analytical principle with the online one except for the sample introduction part. The offline system is used in laboratory to analyze pre-collected aerosol samples on filter substrates. It does not have an inlet system for aerosol sampling. Comparatively, the online nitrogen measurement system is capable of producing high-time resolution data in field campaigns, while the offline system has the advantage of analyzing aerosol samples collected at multiple sites with ease. For nitrogen quantification with the offline system, a filter aliquot is placed in the front oven of the analyzer. During the analysis, the temperature of the front oven is programmed to allow controlled thermal evolution of the aerosol materials into the He/O.sub.2 carrier gas flow. Gaseous products evolved from the front oven are then passed to the back oven, which is packed with a MnO.sub.2 catalyst and held at a preset and programmable temperature (840-870° C.), are oxidized to CO.sub.2 and NO.sub.y. Flow from the back oven is split into two paths, one to a methanator then a flame ionization detector (FID) to measure the carbon amount after converting CO.sub.2 to CH.sub.4, and the other to the molybdenum (Mo) catalytic convertor where NO.sub.y is converted to NO. NO then reacts with ozone to produce excited NO.sub.2*, emitting fluorescence light for quantification. After obtaining C and N thermograms, data treatment process for IN and ON quantification is the same as the description above.

[0046] In accordance with one embodiment of the present invention, the stepwise temperature program is a 6-step temperature program for thermal evolution. Preferably, the 6-step temperature program applies 150° C., 180° C., 300° C., 400° C., 500° C., and 800° C. The carbon signal is monitored by an NDIR detector or a FID which is a part of aerosol carbon analyzer. The nitrogen signals obtained by the chemiluminescent detection are recorded by a data logger at a resolution of 10 Hz (other resolutions such as 100 Hz, 2 Hz, 1 Hz are also suitable, depending on the needs of data processing). The distinct thermal evolution behaviors of inorganic nitrogen and organic nitrogen constituents in ambient samples facilitate the separation of the two nitrogenous aerosols. However, accurate resolution (quantification) of IN and ON relies on MCR data treatment of concurrently measured carbon and nitrogen thermograms showing above. Details of the residence time and evolved nitrogen signal at each heating stage are listed in Table 1.

TABLE-US-00001 TABLE 1 Summary of the major nitrogen signals at each temperature step resolved by the multivariate curve resolution analysis for ambient samples Temperature Time Characteristics of the evolved Step (° C.) (s) nitrogen signal 1 150 500 Dominated by IN: NH.sub.4NO.sub.3 (major), (NH.sub.4).sub.2SO.sub.4 (minor) 2 180 500 Dominated by IN: (NH.sub.4).sub.2SO.sub.4 (major) 3 300 420 Dominated by IN: NH.sub.4HSO.sub.4 (major), (NH.sub.4).sub.2SO.sub.4 (minor) 4 400 420 Both IN and ON 5 500 480 Both IN and ON: ON contributes a larger fraction 6 800 490 ON

[0047] In accordance with one embodiment of the present invention, the method is tested in different standard samples for building calibration curves.

[0048] The calibration curve is constructed by spiking nitrogen standard solutions onto the prebaked filter. The mass concentration of total nitrogen aerosols is quantified based on the calibration curve constructed using single nitrogen standards at five concentration levels in the range of 0.6-3.2 μgN.

[0049] A series of inorganic and organic nitrogenous aerosol standards are selected to verify quantitative determination of nitrogen and to investigate characteristics of the nitrogen thermograms. Major inorganic nitrogen species in the ambient atmosphere, including (NH.sub.4).sub.2SO.sub.4, NH.sub.4HSO.sub.4, and NH.sub.4NO.sub.3, are analyzed. Additionally, organic nitrogen standards with different functional groups are further tested, including three oxidized organic nitrogen compounds (i.e., nitrophenol, 4-nitrocatechol, and nitroguaiacol), two reduced organic nitrogen species (i.e., imidazole and 2-hydroxypyrazine), and a mixture of five single amino acid standards (i.e., alanine, arginine, glycine, histidine, and serine).

[0050] As shown in FIG. 1, the nitrogen standards including an inorganic nitrogen standard (NH.sub.4).sub.2SO.sub.4, an oxidized organic nitrogen standard 4-nitrocatechol, two single reduced organic nitrogen standards: imidazole and 2-hydroxypyrazine and a reduced organic nitrogen mixture standard of amino acid are tested to verify the stability of the presented invention. The response for different inorganic nitrogen and organic nitrogen standards shows excellent consistency. The linear regression yields a nitrogen recovery curve with a slope of 1.02±0.06, and an intercept of 0.01±0.02. For each individual standard tests, the regression exhibits R.sup.2>0.99. This indicates that nitrogen in nitrogenous aerosols with different chemical structures or functional groups can be completely converted to NO.sub.y products and be measured by the present invention.

[0051] In accordance with one embodiment of the present invention, atmospheric organic nitrogen samples and inorganic nitrogen samples are assessed by the present invention.

[0052] Humic-like substances (HULIS) is the hydrophobic fraction of water-soluble organic components in the ambient aerosol. As a mixture of thousands of organic constituents, HULIS extracted from ambient PM.sub.2.5 samples retains the compositional complexity of atmospheric aerosols. The thermogram of HULIS samples can reflect the thermal evolution characteristics of the environmental organic nitrogen mixture.

[0053] FIGS. 2A-2E show the typical nitrogen thermograms under the 6-step temperature program and the corresponding nitrogen masses of the nitrogen standards and ambient samples. According to the thermograms, different nitrogen-containing components exhibit distinct thermal evolution behaviors. Referring to FIGS. 2A-2B, NH.sub.4NO.sub.3 evolves mainly at the first temperature step (150° C.), with minor nitrogen factions coming out in later stages; (NH.sub.4).sub.2SO.sub.4 has nitrogen peaks at 180° C., 300° C., and 400° C.; and NH.sub.4HSO.sub.4 evolves relatively later than (NH.sub.4).sub.2SO.sub.4, starting from 300° C., and produces a significant nitrogen fraction at 400° C. compared with the other two inorganic nitrogen species.

[0054] Referring to FIG. 2C, the atmospheric organic nitrogen complex mixture, HULIS, has N signals throughout all the 6 temperature steps, with very minor contributions from N at the lower temperatures (150 and 180° C.) and a significant N peak at 800° C. This reflects the less volatile nature of aerosol ON mixtures compared with major IN species. FIGS. 2D-2E depict the thermogram and the stepwise nitrogen mass distributions of a whole ambient aerosol sample. As expected, the N thermogram of the whole ambient aerosol sample has peaks at all temperature stages, reflecting combined features of both inorganic and organic nitrogen fractions.

[0055] Referring to FIGS. 3A-3K, example thermograms of individual IN standards (I.e., NH.sub.4NO.sub.3, (NH.sub.4).sub.2SO.sub.4, and NH.sub.4HSO.sub.4), aerosol WSON and WION mixtures and an ambient PM2.5 sample under a 6-step temperature program (150, 180, 300, 400, 500, and 800° C.) (FIG. 3A). Thermal program characteristics of identification utility are described below: (1) NH.sub.4NO.sub.3 produces a major function of N signal at 150° C. and a minor fraction at higher temperature steps (FIG. 3B-3C); (2) (NH.sub.4).sub.2SO.sub.4 evolves mostly at the 180 and 300° C. steps, consistent with a previous mechanistic study reporting its decomposition and releasing NH.sub.3 at 170° C. to form (NH.sub.4).sub.3H(SO.sub.4).sub.2; (3) NH.sub.4HSO.sub.4 produces nearly all N signals at 300° C. steps (FIG. 3D-3E); and (4) aerosol ON mixtures, including both WSON and WION, produce N signals at all temperature steps, but most of their N signals are evolved at the temperature steps of ≥300° C. (FIG. 3F-3I). The minor fractions of WSON and WION appearing at the lowest two temperature steps of 150 and 180° C. are likely associated with the more volatile forms of ON or those nitrogenous materials that are susceptible to facile thermal decomposition. The C thermograms corroborate the organic nature of the WSON and WION mixtures (FIG. 3F and 3H). Note that the thermograms of WSON and WION do not reveal distinguishable features between the two, suggesting that structure moieties underlying water solubility do not drive the thermal evolution behaviors of the ON molecules. The C and N thermograms for the example whole aerosol sample are shown in FIG. 3J, which displays distinctive differences. Specifically, at the lowest two temperature steps (150 and 180° C.), there are significant N peaks but with very little associated C. This reflects that NH.sub.4NO.sub.3, which evolves at 150° C., and (NH.sub.4).sub.2SO.sub.4, part of which evolves at 180° C., are large contributors of aerosol N while only a very minor fraction of organic aerosol materials evolves at these two temperatures. As shown in FIG. 3J, it is noted that over half of the aerosol C contents evolves at the last three temperature steps (400, 500, and 800° C.). These results qualitatively suggest that N signals at 150 and 180° C. steps are mostly ascribed to the evolution of aerosol IN, while aerosol organics evolve dominantly at temperatures higher than 180° C.

[0056] The C/N ratios of the evolved aerosol materials corresponding to the individual temperature steps for 60 ambient PM.sub.2.5 samples analyzed in the embodiments. Note that the carrier gas used here contains oxygen (1% O.sub.2/99% HE) throughout the analysis, which substantially reduces organic carbon pyrolyzation. Table 2 shows that the average C/N ratios for the set of 60 samples at 150 and 180° C. steps are 0.37 and 0.63, respectively, significantly lower than those (6.7-21.7) at the last three higher-temperature steps. As organic nitrogen aerosols contain a C signal while inorganic nitrogen does not contain a C signal, the largely varied C/N ratios at each temperature step qualitatively indicate whether the N signal involved at a given temperature is dominated by IN or likely also contains notable ON. Specifically, the C/N ratio information indicates that the N signals at the lowest two temperatures are largely attributed to IN, while the N signals at temperature steps of ≥300° C. ON are predominantly associated with ON.

TABLE-US-00002 TABLE 2 Mean C/N mass ratios in the evolved aerosol materials at individual temperature steps obtained with 60 ambient samples step Temp (° C.) C/N Remark T1 150 0.37 mainly IN: NH.sub.4NO.sub.3, NH.sub.4Cl T2 180 0.63 mainly IN: NH.sub.4NO.sub.3, (NH.sub.4).sub.2SO.sub.4 T3 300 1.23 NH.sub.4HSO.sub.4, (NH.sub.4).sub.2SO.sub.4, and ON T4 400 6.18 ON T5 500 12.17 ON T6 800 21.66 Mainly ON and KNO.sub.3, NaNO.sub.3

[0057] For quantification of IN and ON by thermograms, multivariate curve resolution (MCR) technique is used. In the aerosol C and N thermograms, while IN and ON signals exhibit distinct thermal characteristics, they are not distinctive peaks. The target of resolving the IN and ON signals is, in principle, similar to the analytical situation of separating overlapping chromatographic peaks with spectroscopic detection resulting from a mixture. A MCR tool has been widely used to solve the overlapping peaks for multicomponent mixture samples. Therefore, MCR technique is applied to resolve signals for IN and ON from their overlapping thermograms. The principle is illustrated in FIG. 5. Specifically, the aerosol IN and ON analyzer generates a data matrix X (n×m), where n is the number of samples and m is the number of measurands (i.e., thermogram peak intensities). The MCR model decomposes the data matrix X into two matrices: factor contributions (C) and factor profiles (P). Mathematically, it can be presented as formula (I):

[00001]$x_{ij}=\underset{k=1}{\overset{p}{.Math.}}{c}_{ik}{p}_{kj}+e_{ij}$

Where x.sub.ij is the measured amount (e.g., concentration) of jth species in the ith sample, c.sub.ik is the contribution of the kth factor to the ith sample, and p.sub.kj is the factor loading of jth species in the kth factor. is the residual amount for each data point. It is clear from the above formula (I) that the underlying principle of MCR is the chemical mass balance. Additionally, a non-negative constrain should be imposed in the application of MCR as a negative factor profile or contribution is physically meaningless for abundance quantities of chemical species. It is noted that positive matrix factorization (PMF), a variant of factor analysis technique widely used to resolve major pollution sources and quantify their contributions, shares the same principle with MCR analysis and has the non-negative constrain incorporated in the model. Thus, the mathematical implementation of the MCR analysis is carried out using USEPA PMF version 5.0. The performance of PMF analysis on predicting TN, IN, and ON masses is tested with laboratory-generated IN+ON mixtures as well as ambient aerosol samples and described in the next subsection. Qualitatively, the thermograms of different N species show distinct variation features, e.g., period of close to zero signal interspersed between discrete peaks in the thermograms, which are desirable for reducing rotational ambiguity in solutions resolved by the multivariate modeling. In other words, the peak areas in the N thermograms and the C thermograms are used as inputs for multivariate curve resolution data treatment to resolve inorganic and organic nitrogen quantities.

[0058] FIG. 4 shows an online atmospheric nitrogen quantification system 100. Because the system delivers results as samples are collected, results can be posted online for publication of conditions in a particular area. System 100 includes an inlet 101. In one embodiment, inlet 101 may a sharp-cut cyclone with an aerodynamic cut-off diameter of 2.5 μm and ambient air can be drawn through the inlet at a rate of 8.0 L/min for sampling ambient air. The sampled air passes through a parallel-plate filter denuder 102 for reducing positive artifacts caused by adsorption of gas-phase compounds and then into a carbon analyzer 103 having a front oven 104 for carrying out a stepwise temperature program and thermally evolving the aerosol sample collected onto the quartz filter and a back oven 105 for oxidizing the gaseous product with an oxygen-containing carrier gas, such as a He/O.sub.2 carrier gas.

[0059] After the carbon analyzer 103, the gas is split into two paths. One enters the NDIR detector 106 for carbon amount quantification. The other passes through a flow-adjustable needle valve 107 and enters a NO.sub.x analyzer 108 having a molybdenum catalytic convertor where it is transformed into NO. The transformed gas reacts with ozone, which is generated by an ozonator 109, and produces excited NO.sub.2* molecules in a reaction chamber 110. A fluorescence quantification module 111 utilizes a photomultiplier tube to evaluate and quantify the intensity of fluorescence emitted during the chemiluminescent reaction. A data logging module 112 analyzes, calculates and saves the fluorescence intensity data. The intensity value of the fluorescence represents a total nitrogen mass concentration of the ambient air. Since this intensity value is generated in real time, the results are posted online for real time updates of environmental conditions.

[0060] The apparatus additionally includes a temperature control module 113 for adjusting and controlling the stepwise temperature program of the front oven 104 and the temperature of the back oven 105, and a controller 114 to control the data logging module 112 and the temperature control module 113. In one embodiment, the front oven has a pre-baked filter to filter the gas sample to obtain an aerosol product for further thermally evolving.

EXAMPLES

Example 1

[0061] The present invention is used to evaluate ambient inorganic and organic nitrogen aerosols. The testing sampling is conducted at the Hong Kong University of Science and Technology (HKUST) site. The statistical summary of the ambient sample is listed in Table 3.

TABLE-US-00003 TABLE 3 Statistical summary (average ± standard deviation) of meteorological parameters, air quality data, and aerosol nitrogen mass concentrations at the HKUST site during the measurement period (29 Nov. 2020-1 May 2021). Subperiod Overall Winter Spring Date 29 Nov. 2020- 29 Nov. 2020- 1 Mar. 2021- 1 May 2021 28 Feb. 2021 1 May 2021 Dominant WD — North Southeast RH (%) 74.1 ± 17.5 69.3 ± 17.4 85.4 ± 11.3 T (° C.) 18.3 ± 4.2  16.9 ± 3.8  21.7 ± 2.7  WS (m/s) 3.0 ± 1.7 3.1 ± 1.8 2.7 ± 1.3 O.sub.3 (ppb) 45.4 ± 17.7 44.9 ± 16.4 47.0 ± 21.1 SO.sub.2 (ppb) 1.0 ± 0.7 1.0 ± 0.8 1.2 ± 0.6 CO (ppb).sup.a 605 ± 186 686 ± 133 415 ± 152 NO.sub.x (ppb) 7.9 ± 6.8 7.8 ± 6.8 8.1 ± 6.9 PM.sub.2.5 (μg/m.sup.3).sup.a 16.7 ± 7.8  17.4 ± 7.8  15.0 ± 7.5  TN (μgN/m.sup.3) 1.84 ± 0.99 2.01 ± 1.04 1.41 ± 0.65 ON (μgN/m.sup.3) 0.24 ± 0.09 0.26 ± 0.09 0.22 ± 0.08 IN (μgN/m.sup.3) 1.59 ± 0.93 1.76 ± 0.99 1.19 ± 0.60

[0062] The measurement period spanned two seasons. In the winter subperiod (November-February), the prevailing winds were northerly and northeasterly while in the spring (March-May) subperiod, the southeast wind dominated. The PM.sub.2.5 concentration level is lower in spring than that in winter (15.0 vs. 17.4 μg/m.sup.3).

[0063] The online aerosol nitrogen quantification method quantifies the atmospheric abundances of aerosol total nitrogen, inorganic nitrogen and organic nitrogen. FIG. 6 shows the time series of the nitrogenous aerosol concentrations, along with the NO.sub.x and PM.sub.2.5 concentrations. During the measurement period, hourly aerosol total nitrogen (TN) is in a range of 0.18-6.33 μgN/m.sup.3 with an average of 1.84±0.99 μgN/m.sup.3. Inorganic nitrogen (IN) concentration is 0.07-5.89 μgN/m.sup.3 with an average of 1.59±0.93 μgN/m.sup.3. The concentration of organic nitrogen (ON) is 0.04-0.67 μgN/m.sup.3 with an average of 0.24±0.09 μgN/m.sup.3. Both inorganic nitrogen and organic nitrogen show lower concentrations in spring compared with those in winter (IN: −32.4%; ON: −15.4%). This is mainly associated with seasonal variations in strength of different nitrogen sources, weather conditions, and background concentrations which are largely influenced by regional transport of air pollutants. Overall, ON accounted for 13.0% of TN, indicating the significant presence of ON species in ambient nitrogenous aerosols. The ON fraction in TN is elevated in spring (15.6%) compared with winter (12.9%).

Example 2

[0064] To investigate the influential factors of aerosol ON formation and to demonstrate the utility of online measurements of ON, three short-term high ON episodes are examined, as presented in FIGS. 7A-7C. As shown in FIG. 7A, the first high ON episode occurred during 16:00-20:00 on the Christmas Eve (24 Dec. 2020). Driven by the increased traffic flow on holiday, a surge in NO.sub.x level occurred within hours along with rapid increase in both aerosol IN and ON. During the episode, the wind changed from eastern to southern and southwestern direction, which led to a transition from the relatively clean coastal air masses to the air of higher pollution loading from the nearby urban areas. In the first episodic period, ON is significantly correlated with secondary inorganic nitrogen components, NH.sub.4.sup.+—N and NO.sub.3.sup.−—N (R.sub.p=0.94 and 0.85, respectively) as well as three SOA tracers associated with isoprene, α-pinene, and β-caryophyllene emissions (R.sub.p=0.90, 0.91, and 0.96, respectively), indicating the prominent influence from secondary formation processes. As demonstrated by Perring et al. (2013), secondary formation serves as an important source of atmospheric ON aerosols. Similar important contributions from secondary formation processes to the WSON have been observed in northwestern and southern China (Ho et al., 2015; Yu et al., 2017).

[0065] In the second case period of high ON episode (FIG. 7B), the dominant wind shifted from southeastern and southern to northern direction with increased wind speed (from <1 m/s to >5 m/s) during the midnight of 29-30 Dec. 2020, bringing air masses from the continental area of mainland China. This episode shows a clear example of nighttime nitrogen production under the influence of biomass burning and secondary formation processes, as suggested by the concomitant rise and fall of ON, levoglucosan, 4-nitrocatechol, 4-nitrophenol, isoprene SOA tracers, and α-pinene SOA tracers.

[0066] As shown in FIG. 7C, the third ON episodic case lasted from the late evening of 17 February to the noon of 18 Feb. 2021. Different from the first two experiments with increases of both aerosol IN and ON under high NO.sub.x levels (maximum=78 and 41 ppb, respectively), the third case period is characterized with pronounced ON increment in relatively low NO.sub.x regime (<8 ppb). As reflected by the correlation analysis, biomass burning emissions represent the major source of aerosol ON in the third field experiment (R.sub.p=0.80 and 0.94 for ON versus levoglucosan and mannosan, respectively), while contributions from the biogenic SOA formation processes are relatively weak.

[0067] FIGS. 8A-8D present the scatter plots of aerosol ON versus NO.sub.x, IN, and two IN species (i.e., NH.sub.4.sup.+—N and NO.sub.3.sup.−—N) for the entire measurement periods with the data points from the three episodes highlighted. Considering all the data, there are broad positive correlations between ON and all the associated N species, especially with IN and NH.sub.4.sup.+—N (FIG. 8C) showing stronger correlations with ON than NO.sub.x (FIG. 8A) and NO.sub.3.sup.−—N (FIG. 8D) do. Such results reflect varied precursor-source connections over the five-month observation period. Expectedly, much higher correlations of ON with NH.sub.4.sup.+—N , IN, and NO.sub.x are observed for the short periods covering the individual cases due to the more uniform underlying chemical and physical conditions. However, during the third ON episodic case period, there is nearly no correlation between NO.sub.3.sup.−—N and ON, showing a case of decoupling of the formation/sources of ON and NO.sub.3.sup.−. This demonstrates the values of having high-time resolution of ON measurements. As shown in FIG. 8B, the scatter plots of ON vs IN among all the three case periods also reveal that the relative importance of ON and IN significantly varies under different pollution conditions. The regression slopes of the three episodic cases range from the smallest value of 0.05 in the first case to the largest of 0.27 in the third case. Such results imply that although aerosol IN and ON have certain precursors in common, such as NO.sub.x and NH.sub.3, their variations in atmospheric abundance can still be largely different due to the complex primary and secondary source origins of ON and distinct IN and ON formation mechanisms.

[0068] Measurement Validation

[0069] The present nitrogenous aerosol quantification method is further validated through comparison with the IN measured by the traditional offline analysis with ion chromatography (IC).

[0070] During the ambient measurement at HKUST, 6-h filter samples are collected on 47 mm Teflon filters and prebaked quartz filters by using different channels of a Speciation Air Sampling System (SASS, Met ONE Instruments Inc., USA) at the sampling flow rate of 6.7 L/min. The offline PM.sub.2.5 filter samples are extracted by the double de-ionized water in an ultrasonic ice-water bath and filtered through a 0.45 μm syringe filter. Water-soluble ions are analyzed using an IC system (Dionex DX-500, Thermo Fisher Scientific, USA) following the standard procedure known in the art.

[0071] A total of 15 groups of 6-h data points are obtained by the offline quartz filters and the offline Teflon filters. The hourly IN data of the online IN & ON analyzer system in the corresponding time periods are averaged for comparison with the offline data. The aerosol IN filtered by the quartz and Teflon filters are highly correlated (R.sup.2=0.96), with a slope value of 0.89±0.05 (FIG. 9C). As shown in FIGS. 9A and 9B, the aerosol IN determined by the online method is strongly correlated with those determined by the offline method (R.sup.2=0.98, slope=0.97 for samples collected on quartz filters; R.sup.2=0.96, slope=1.05 for samples collected on Teflon filters).

[0072] The regression slopes for IN measured online versus offline with filters fall into a close range around 1.0, demonstrating the good performance of aerosol IN quantification by the present invention. There are several reasons for the discrepancies between the two methods: (1) a parallel-plate filter denuder is installed upstream of the online atmosphere nitrogen quantification system to minimize the interference from semi-volatile species (e.g., NH.sub.4NO.sub.3). In contrast, the SASS system sampled directly without a denuder, consequently introducing the positive adsorption artifacts; (2) filter face velocity, defined as the volumetric flow rate divided by the exposed area of the filter, influences the vapor adsorption onto the filter medium. A previous study demonstrated the face velocity effects on vapor adsorption artifact in the sampling of organic aerosols (McDow and Huntzicker, 1990). IN measurements may also have similar face velocity dependence caused by adsorption of gas-phase HNO.sub.3 on the quartz filter. With the similar volumetric flow rate (8.0 vs. 6.7 L/min) but much smaller filter deposit area (1.23 cm.sup.2 vs. 13.8 cm.sup.2), the online system has a filter face velocity ˜13 times that of the offline sampling system, leading to less interference from gases. Such characteristics could be even more critical for ON measurement; (3) the filter-based sampling followed by offline laboratory analysis experiences a longer period of gas-particle partitioning and chemical equilibrium during the storage and conditioning process; and (4) the inexact matching of the sampling schedules between the offline (continuous 6-h) and the online (the actual sampling time is 3 h in the 6-h period) sampling may also contribute to the discrepancies in IN determination.

[0073] The foregoing description of the present invention has been provided for the purposes of illustration and description. It is not intended to be exhaustive or to limit the invention to the precise forms disclosed. Many modifications and variations will be apparent to the practitioner skilled in the art.

[0074] The embodiments were chosen and described in order to best explain the principles of the invention and its practical application, thereby enabling others skilled in the art to understand the invention for various embodiments and with various modifications that are suited to the particular use contemplated.