1 Introduction

For decades, China has experienced frequent and all too often severe hazy weather, especially in the Beijing-Tianjin-Hebei region, the Pearl River Delta, the Yangtze River Delta and the areas in and around Chengdu and Chongqing (Sun 2016). Light extinction caused by aerosol particles dispersed in the atmosphere leads to reductions in atmospheric visibility that are characteristic of the most common types of hazy weather (He et al. 2013). Moreover, the emission sources for aerosols are numerous, and the composition of aerosols is complex, and these complicating factors have serious implications for the ecological environment, global climate, and human health (Chen and Xiao 2009; Shao et al. 2018; Zhao et al. 2019). As a result, atmospheric aerosols have become a hot topic for domestic research in China and throughout the world (He et al. 2013; Huang et al. 2014).

Advances in industrial desulfurization technology and the strict monitoring of sulfur dioxide (SO2) in China have led to some improvements in air quality. However, increases in nitrogen oxide (NOx) emissions caused by human activities (Wen 2015) have led to a gradual increase in nitrate (NO3−) concentrations. Ammonium (NH4+) is mainly produced by the conversion of ammonia gas (NH3) in the troposphere, which is the third most abundant nitrogen-containing gas in the atmosphere after N2 and N2O, and together with NH4+, it is collectively referred to as NHx (NH3 + NH4+) (Seinfeld et al. 1998; Walters and Hastings 2018). NO3− and NH4+ are important components of secondary inorganic aerosols (SNA), and they participate in various reactions that affect the global nitrogen cycle (Behera et al. 2013; Tang et al. 2006; Walters and Hastings 2018). Recent studies have shown that precursors of NO3− also can promote the formation of sulfates and in so doing aggravate atmospheric pollution (Cheng et al. 2016). Therefore, it is of great importance to understand the chemical formation of NO3− and NH4+ and how these substances are removed from the atmosphere.

Moreover, the sources and mechanisms leading to the formation of NO3− and NH4+ in urban aerosols are not yet fully understood, and the rapid increases in concentrations of NO3− during heavy pollution is also uncertain. Stable isotopes have become more widely used in studies of the sources, chemical formation and migration process of pollutants in recent years (Zhao et al. 2019), and the determination of the N and O isotopic composition of NO3− can be used to identify sources and investigate conversion processes (Xu et al. 2014; Zhao et al. 2019). The application of stable isotope techniques holds promise for tracing the movements of atmospheric NO3−, and similarly, N isotope measurements of NH4+ may provide insights into N sources and removal processes (Walters and Hastings 2018). Therefore, this paper reviews the existing research on N isotope analysis of NO3− and NH4+.

2 Analytical Techniques for Nitrogen Isotopes in Nitrate and Nitrogen Oxides

2.1 Nitrogen Isotopes and the Formation of Nitrogen Oxides

NO3− in the atmosphere mainly forms through reactions of NOx and free radicals. It is generally agreed that during the daytime, the formation of NO3− is mainly driven by gas-phase reactions between NO2 and OH radicals (NO2 + OH) (Fig. 1) (Seinfeld and Pandis 2016; Tang et al. 2006). At night, HNO3 mainly forms through reactions of N2O5 in the liquid phase (N2O5 + H2O) (Fig. 2) (Jia et al. 2006). The factors that affect the formation of NO3− include not only temperature but also relative humidity (RH) and solar radiation intensity (Elliott et al. 2009). In addition, N2O5 at night also forms through multi-step reactions of NOx involving O3. The main sources for gaseous NOx include emissions from soil microbes, biomass burning, thermal power plants and motor vehicle exhaust (Bao et al. 2015; Wang et al. 2014).

Fig. 1
figure 1

Oxidation of NOx to NO3− during the day (Seinfeld and Pandis 2016; Tang et al. 2006)

Full size image
Fig. 2
figure 2

Oxidation of NOx to NO3− at night (Jia et al. 2006)

Full size image

Previous studies have shown that the normalized ratios of 14N to 15N isotopes (δ15N) of NOx to its oxidation product, NO3−, can be used as a type of "fingerprint" to evaluate the contributions of NOx from various sources to atmospheric nitrogen deposition (Elliott et al. 2007). The N isotope ratios for NOx vary among sources, and therefore, the stable N isotope ratio for NO3− (δ15N-NO3−) can be used to trace its main sources. In addition, 15N fractionation occurs during the oxidation of NOx to NO3− (Walters and Michalski 2016), and therefore, N isotopes also can provide insights into NO3− formation (Walters and Michalski 2016; Zhang and Gao 2012; Zheng 2000). A limitation in the application of N isotope for source assessments is that the δ15N values for NOx from different sources can overlap to a degree, and the fractionation that occurs when NOx is converted to NO3− can be another confounding effect. To maximize the usefulness of isotopic information, both N and O isotopes are often measured together in the stable isotope analysis of NO3− (Qin et al. 2019). This approach makes use of the fact that the oxygen isotope composition of O3 and OH radicals are different (δ18O-O3: + 90 to + 120‰ versus δ18O-OH: − 89 to − 56‰) (Johnston and Thiemens 1997; Michalski et al. 2012). Indeed, δ18O-NO3− can provide additional insights into the formation of NO3− because data for both O and N isotopes can compensate for the limitations in relying on NO3− nitrogen isotopes alone (Altieri et al. 2013; Wankel et al. 2009). Therefore, the use of isotope techniques can effectively investigate the sources for NO3− and the oxidation pathways for NOx.

2.2 Methods for Measuring Nitrogen Isotopes in Nitrate

Quartz fiber filters are commonly used to collect aerosol NO3− particles, and then, ultrapure water is used to extract NO3− for subsequent analysis. The solution is then used for the N and O isotope determination of NO3− (Zhao et al. 2019). At present, the methods for determining the N isotopes in NO3− involve distillation, diffusion, graphite combustion, ion exchange resins, denitrifying bacteria, and chemical reduction (Qin et al. 2019; Zhao et al. 2019).

The operating principles for the distillation and diffusion methods are straightforward; that is, NO3− is reduced to NH4+ for the instrumental analysis. However, these two methods require relatively large sample sizes, and isotope fractionation is likely to occur during the procedures, resulting in large uncertainties in the results. More important, these techniques cannot be used to determine nitrogen and oxygen isotopes simultaneously, so their applications are limited (Liu et al. 2017; Xiao and Liu 2001b; Zhao et al. 2019). For the graphite combustion procedure, which has been improved several times, the NO3− sample and granular graphite are first placed in a glass/graphite carbon reaction tube, and then they are converted into N2 and CO at a high pyrolysis temperature of 1400 °C. After purification, N and O isotopes can be simultaneously determined by mass spectrometry, and this method been automated (Kornexl et al. 1999; Qin et al. 2019). The ion exchange method uses an anion resin to adsorb and concentrate NO3− from aqueous solution, and HCl is then used to remove the adsorbed NO3−. After a series of purification steps, N and O isotopes can be determined by mass spectrometry (Qin et al. 2019; Silva et al. 2000). However, the pretreatment steps for the ion exchange method are cumbersome, require large sample quantities, and the cost of pretreatment is high because of the consumables used; therefore, this method is not often used (Liu et al. 2017; Xiao and Liu 2001b).

More recently, denitrifying bacteria and chemical reduction methods have developed for isotopic NO3− analyses with N2O as the final product. These methods are not sensitive to atmospheric background, and they can be used for small samples (usually 10–60 nmol) (Casciotti et al. 2002; Mcilvin and Altabet 2005; Sigman et al. 2001). The molecular structure of N2O is such that the δ15N value for this compound is the average of the δ15N for the two nitrogen atoms (Wang et al. 2015) as follows:

$$\delta^{15} {\text{N}}_{{\left( {{\text{N}}_{{2}} {\text{O}}} \right)}} = \frac{{\delta^{15} {\text{N}}_{{\left( {{}_{ }^{{{15}}} {\text{N--N--O}}} \right)}} + \delta {}_{ }^{15} {\text{N}}_{{\left( {{\text{N}-\text{}}{}_{}^{{{15}}} {\text{N} -\text{O}}} \right)}} }}{2}.$$
(1)

The denitrifying bacteria method uses obligate denitrifying bacteria lacking N2O reductase to reduce all NO3− to N2O in the sample solution. Nitrogen and oxygen isotopes are then determined using a mass spectrometer (Casciotti et al. 2002; Sigman et al. 2001; Yang et al. 2014; Zhang et al. 2010). The denitrifying bacteria method is suitable for different types of samples, and it can be used to measure very low concentration (10–20 nmol N) samples (Liu et al. 2017; Sigman et al. 2001). Meanwhile, the precision of the final analytical results is very high (around 0.2‰) (Sigman et al. 2001). The drawbacks for the method are complex biological responses, poor universality, degradation of bacteria characteristics, and high maintenance costs, and therefore, it is not so commonly used (Wang et al. 2015; Zhao et al. 2019). A chemical reduction method has been developed by Mcilvin and Altabet (2005), and it is based on two reduction steps. The first step is to reduce NO3− to NO2− over cadmium (Cd), and the second step is to use the sodium azide (NaN3) prepared in an acidic buffer (commonly an acetic acid buffer) to quantitatively reduce NO2− to N2O. After that, the ratios of N and O isotopes can be determined by isotope ratio mass spectrometry. The specific reaction equation is as follows (Wang et al. 2015):

The first step of the reduction method occurs in a weakly alkaline environment, and Cd is used to reduce NO3− in the solution to NO2−:

$${\text{NO}}_{3}^{ - } + {\text{Cd}} + {\text{H}}_{2} {\text{O}} \to {\text{NO}}_{2}^{ - } + {\text{Cd}}^{2 + } + 2{\text{OH}}^{ - } .$$
(2)

The second step of the reduction method is in a sodium azide-acetic acid buffer system, where N3− is used to reduce NO2− to N2O gas:

$${\text{NO}}_{2}^{ - } + {\text{H}}^{ + } \to {\text{HNO}}_{2} ,$$
(3)
$${\text{HNO}}_{2} + {\text{H}}_{2} {\text{O}} \to {\text{H}}_{2} {\text{NO}}_{2}^{ + } + {\text{OH}}^{ - } ,$$
(4)
$${\text{H}}_{2} {\text{NO}}_{2}^{ + } + {\text{N}}_{3}^{ - } \mathop \to \limits^{{{\text{slow}}}} {\text{ N}}_{3} {\text{NO}} + {\text{H}}_{2} {\text{O}} \mathop \to \limits^{{{\text{fast}}}} {\text{ N}}_{2} {\text{O}} + {\text{N}}_{2} .$$
(5)

The chemical reduction method has a wide range of applications and is operationally simple, highly sensitive, and relatively inexpensive, and it is now commonly used in China (Liu et al. 2017; Tu et al. 2016; Wang et al. 2015). However, a shortcoming of this method is that the azide reagent is toxic, requiring careful operation in a fume hood and appropriate safety measures in cleaning operations (Liu et al. 2017; Mcilvin and Altabet 2005).

2.3 Nitrogen Isotope Determination Methods for Nitrogen Oxides

The methods for determining N isotopes in NOx are similar to those used for NO3−. The biggest difference between them is in the methods used for collection and pretreatment. For the collection of NOx, a gas absorption bottle is typically used (Felix et al. 2012; Wang et al. 2015). After the end of sampling, the sealed gas absorption bottle, which contains a fixed volume of an absorption liquid, is shaken to oxidize the NOx to NO3−, the later of which becomes dissolved in the absorbing liquid. The bottle is left for a sufficient amount of time to ensure that all NOx is converted to NO3−. The reaction equation for the oxidation of NOx to NO3− by H2O2 is as follows (Wang et al. 2015):

$${\text{NO}}_{2} + {\text{NO}} + 3{\text{H}}_{2} {\text{O}}_{2} \mathop \to \limits^{{{\text{H}}^{ + } }} { }2{\text{NO}}_{3}^{ - } + 3{\text{H}}_{2} {\text{O}}{.}$$
(6)

The methods used for the determination of NOx nitrogen isotopes are the same as those previously described for NO3− nitrogen isotopes in Sect. 2.2.

2.4 Recent Research Progress

Tu et al. (2016) improved the chemical reduction methods (cadmium + hydrazoic acid) that were developed for determining N and O isotopes in NO3−. First, the sponge cadmium or activated cadmium was replaced with cadmium powder for the NO3− reduction step, and the reaction was carried out under ultrasound. Second, the amount of sodium azide in the acetic acid buffer was decreased to only 1% of the original method for the second reduction step; that reduced the amount of toxic reagents used. Third, sodium acetate was added to the buffer along with the sodium azide because that improved the performance of the buffer and decreased the oxygen isotope exchange with water during the NO2− reduction process (Liu et al. 2017).

The types of cadmium powders used for the chemical reduction methods can affect the reaction systems and kinetics, pH, etc. (Tu et al. 2016; Wang et al. 2015). Zhao et al. (2019) studied the two-step chemical reduction methods under different conditions and optimized the scheme for accurate determination of N and O isotopes in NO3− aerosols. Their studies showed that when the concentration of Cl− in the solution was ≥ 5 mol/L and pH 8, the cadmium powder was activated by HCl and that favored the reduction of NO3− to NO2−. And then, NO2− was reduced to N2O with the sodium azide in an acetic acid buffer at a pH of 4.5–4.6, and isotopic determinations were performed with an isotope ratio mass spectrometer. The measurement accuracy for N isotopes determined by this method is ± 0.08%, and the corresponding accuracy for O isotopes is ± 0.24%.

3 Nitrogen isotope Analytical Techniques for Ammonium and Ammonia

3.1 Formation Mechanisms for Ammonium and Tracing the Reactions with Nitrogen Isotopes

Ammonium (NH4+) ions are formed by the reaction of gaseous NH3 with acids in the atmosphere, mainly H2SO4 and HNO3. Reaction products include ammonium hydrogen sulfate (NH4HSO4), ammonium sulfate ((NH4)2SO4), and ammonium nitrate (NH4NO3), all of which are important aerosol particle components that can degrade atmospheric visibility (Liu 2017). H2SO4 and HNO3 are most often formed through the oxidation of SO2 and NO2 in the atmosphere by OH radicals, O3, H2O2, etc., and they further react with NH3 to form NH4HSO4, (NH4)2SO4 and NH4NO3 (Kong et al. 2014; Wang et al. 2012). The specific reaction equations are as follows (Seinfeld and Pandis 2016; Wang et al. 2009):

  1. 1.

    Formation of H2SO4:

    $${\text{SO}}_{2} + {\text{OH}} \cdot + {\text{M}} \to {\text{HOSO}}_{2} \cdot + {\text{M,}}$$
    (7)
    $${\text{HOSO}}_{2} \cdot + {\text{O}}_{2} \to {\text{HO}}_{2} \cdot + {\text{SO}}_{3} ,$$
    (8)
    $$\begin{gathered} {\text{SO}}_{3} + {\text{M}} + {\text{H}}_{2} {\text{O}} \to {\text{H}}_{2} {\text{SO}}_{4} + {\text{M,}} \hfill \\ {\text{Or:}} \hfill \\ \end{gathered}$$
    (9)
    $${\text{SO}}_{2} + {\text{H}}_{2} {\text{O}} \Leftrightarrow {\text{H}}_{2} {\text{SO}}_{3} ,$$
    (10)
    $$3{\text{H}}_{2} {\text{SO}}_{3} + {\text{O}}_{3} \to 3{\text{H}}_{2} {\text{SO}}_{4} ,$$
    (11)
    $${\text{H}}_{2} {\text{SO}}_{3} + {\text{H}}_{2} {\text{O}}_{2} \to {\text{H}}_{2} {\text{SO}}_{4} + {\text{H}}_{2} {\text{O}}{.}$$
    (12)
  2. 2.

    Formation of HNO3:

    $$\left\{ {\begin{array}{*{20}c} { {\text{NO}}_{2} + {\text{OH}} \cdot + {\text{M}} \to {\text{HNO}}_{3} + {\text{M}} \left( {{\text{day}}\,{\text{time}}} \right)} \\ {{\text{N}}_{2} {\text{O}}_{5} + {\text{H}}_{2} {\text{O}} \to {\text{HNO}}_{{3{ }}} \, \left( {{\text{night}}\;{\text{time}}} \right)} \\ \end{array} } \right..$$
    (13)
  3. 3.

    Formation of NH4+:

    $${\text{NH}}_{3} + {\text{H}}_{2} {\text{SO}}_{4} \to {\text{NH}}_{4} {\text{HSO}}_{4} ,$$
    (14)
    $${\text{NH}}_{3} + {\text{SO}}_{3} + {\text{H}}_{2} {\text{O}} \to {\text{NH}}_{4} {\text{HSO}}_{4} ,$$
    (15)
    $${\text{NH}}_{3} + {\text{NH}}_{4} {\text{HSO}}_{4} \to \left( {{\text{NH}}_{4} } \right)_{2} {\text{SO}}_{4} ,$$
    (16)
    $$2{\text{NH}}_{3} + {\text{H}}_{2} {\text{SO}}_{4} \to \left( {{\text{NH}}_{4} } \right)_{2} {\text{SO}}_{4} ,$$
    (17)
    $${\text{NH}}_{3} + {\text{HNO}}_{3} \to {\text{NH}}_{4} {\text{NO}}_{3} .$$
    (18)

H2SO4 has a lower saturation vapor pressure than HNO3, and therefore, H2SO4 will preferentially react with NH3 to form NH4HSO4 and (NH4)2SO4 (Baek and Aneja 2004; Liu 2017). That is, when the concentration of NH3 is low, it will preferentially react with H2SO4 to form NH4HSO4 and (NH4)2SO4, but when NH3 is in abundance, the excess NH3 will react with HNO3 to form NH4NO3. The main factors affecting the formation of NH4+ are the concentrations of NH3, SO2, NOx and other precursors, temperature, relative humidity (RH), solar light intensity and weather conditions (Hu 2015; Liu 2017). In particular, the main factors affecting the reversible reaction between NH3 and HNO3 that forms NH4NO3, are the concentration of NH3, and temperature (Hueglin et al. 2005). In addition, NH3 is the main precursor of NH4+, and it can be oxidized by OH radicals to form NOx, which in turn is involved in the formation of NO3− in a related nitrogen cycle (as shown in Fig. 1) (Seinfeld and Pandis 2016; Tang et al. 2006).

The source of NH3 in different regions is generally different. In agricultural areas, NH3 is mainly derived from animal husbandry and the use of fertilizers (Zbieranowski and Aherne 2013). In urban areas, NH3 has many sources, including industrial sources, vehicle exhaust, external area transmission, solid waste, sewage, and human feces (Dentener and Crutzen 1994; Dong et al. 2010; Gu et al. 2012). In the suburbs, NH3 is mainly derived from natural emissions such as microbial activities and animal wastes (Sutton et al. 2009). The nitrogen isotope composition of NH3 (δ15N-NH3) emitted from different emission sources is generally different (Fig. 3), and the nitrogen isotope composition of NH4+ (δ15N-NH4+) produced is also different. Therefore, the nitrogen isotope composition of NH4+ in atmospheric aerosols can effectively trace its source and formation and transformation processes (Felix et al. 2013).

Fig. 3
figure 3

Previously reported δ15N-NH3 values of NH3 source emissions (Chang et al. 2016; Felix et al. 2013; Freyer 1978; Heaton 1987; Schulz et al. 2001; Smirnoff et al. 2012)

Full size image

Previous studies have used a wide variety of collection techniques for measuring the nitrogen isotope composition of NH3 emission sources (δ15N-NH3), and it has been found that there can be large variations in δ15N from a single NH3 emission source. There also can be considerable overlap in the δ15N values for different NH3 emission sources (Chang et al. 2016; Felix et al. 2013; Freyer 1978; Heaton 1987; Hristov et al. 2009; Savard et al. 2017; Smirnoff et al. 2012; Walters and Hastings 2018). This suggests that there needs to be better information on nitrogen isotopes for that major NH3 sources, and the development of a representative database will require extensive field sampling with many actual measurements (Zong et al. 2017). Another complicating factor is that some processes involving NH3 can cause isotope fraction: one example of this is the evaporation of NH3 which can lead to increases in δ15N in the sample.

In addition, when NH3 is converted to NH4+, equilibrium fractionation of the N isotopes occurs (Freyer 1978; Savard et al. 2017), and that affects the isotopic composition of NH4+ and NH3. The specific reaction equation is as follows (Urey 1947):

$${}_{ }^{15} {\text{NH}}_{3} + {}_{ }^{14} {\text{NH}}^{4 + } \rightleftharpoons {}_{ }^{14} {\text{NH}}_{3} + {}_{ }^{15} {\text{NH}}^{4 + } .$$
(19)

The equilibrium constant (K) or isotope fractionation coefficient (α) for the above formula is 1.034 ± 0.002 at 298.1 K (Urey 1947), and therefore

$$K_{{298.1\;{\text{K}}}} = \alpha_{{298.1\;{\text{K}}}} = \frac{{{}_{ }^{15} {\text{NH}}^{4 + } /{}_{ }^{14} {\text{NH}}^{4 + } }}{{{}_{ }^{15} {\text{NH}}_{3} /{}_{ }^{14} {\text{NH}}_{3} }} = 1.034 \pm 0.002.$$
(20)

This shows that conversions between NH3 and NH4+ should favor the partitioning of 15N into NH4+, and the net result would be higher δ15N values for NH4+ compared with NH3 (Urey 1947). Therefore, the determination of δ15N for NHx provides information on the heterogeneous conversion of NH3 into new particles, and that has implications for the relative importance of kinetics versus equilibrium control in the formation of NH4+. As the δ15N of NH3 may change during the formation of new particles, the form of NHx must be considered in field sampling and laboratory measurements, and the source of the emissions should also be determined to minimize the effects of chemical and physical processes and to obtain meaningful data for the δ15N in NH4+ and NH3 (Walters and Hastings 2018).

3.2 Nitrogen Isotope Determination Methods for Ammonium Ion

Quartz fiber filters are generally used for the collection of aerosol NH4+, and ultrapure water is then used to extract the NH4+ from the particles, and the solution is then used for the nitrogen and oxygen isotope determinations of NH4+.(Huang et al. 2018; Xiang et al. 2019). The methods currently used for determining NH4+ nitrogen isotopes (δ15N-NH4+) can be divided into two categories: Duma and chemical conversion methods (Liu et al. 2017). The Duma methods are traditional, mainly including distillation, diffusion, ion exchange, etc. These methods usually convert the NH4+ extracted from the sample into N2 by combustion and oxidation, and then the N isotopes of the NH4+ are determined by isotope ratio mass spectrometry (IRMS) (Liu et al. 2017; Wen et al. 2016; Xiang et al. 2019).

For the distillation method, the filter extract is heat-distilled in an alkaline environment to convert NH4+ into NH3, which is then absorbed by an acidic liquid (most often H2SO4) (Freyer 1978; Heaton 1987; Velinsky et al. 1989). Then the Dumas combustion method is used to convert NH4+ into N2 for the measurements of nitrogen isotopes (Liu et al. 2017; Wen et al. 2016). Although this method has a high nitrogen recovery rate, it is prone to cross-contamination of 15N. The diffusion method for NH4+ nitrogen isotope measurements is similar to the distillation method, but for the diffusion method, the separated NH3 is absorbed into a solution using a filter device (Holmes et al. 1998). The diffusion method is less prone to cross-contamination, but the absorption time is long (Zeng et al. 2013). The ion exchange method for NH4+ nitrogen isotopes uses a resin to adsorb NH4+, and after loading, the adsorbed ions are removed with a suitable eluent (Xiao and Liu 2001a, 2002). Unfortunately, the adsorption of NH4+ in this method is susceptible to artifacts from strong ions (Hu and Liu 2013). In addition, all of the Duma methods use N2 for the IRMS analysis making them susceptible to interferences from the atmospheric N2 background. As a result, these methods are not suitable for the determination of low concentration samples.

Zhang et al. (2007) developed a new analytical technique for low-concentration samples (< 0.5 µM) in which NH4+ is chemically converted to N2O prior to mass spectrometry. The first step for the procedure is to add a low concentration of BrO− solution to the original acidic absorption solution. The NH4+ is oxidized to NO2− which in turn is converted into N2O for N isotope determinations with the use of sodium azide in an acetic acid buffer or hydroxylamine hydrochloride (Liu et al. 2014; Zhang et al. 2007). However, as noted above, hydrazoic acid is highly toxic, and it is also volatile, so it must be handled carefully in a fume hood. In comparison, hydroxylamine hydrochloride (NH2OH·HCl) solutions are less volatile and less toxic. Indeed, Liu et al. (2014) used hydroxylamine hydrochloride to reduce the NO2− which is from the first step to N2O and then determined the N isotopes in NH4+. This recently developed conversion method does not require the use of toxic reagents, and it also has the advantage of being simple operationally. The precision of the final results is also very high (~ 0.3‰), and as a result, it has recently received much attention. The specific reaction equations are as follows:

  1. 1.

    In an alkaline environment, BrO− oxidizes NH4+ to NO2− (Zhang et al. 2007):

    $${\text{BrO}}^{ - } + {\text{NH}}_{4}^{ + } \to {\text{NO}}_{2}^{ - } + 3{\text{H}}_{2} {\text{O}} + 3{\text{Br}}{.}$$
    (21)
  2. 2.

    In an acidic environment, hydroxylamine hydrochloride reduces the NO2− to N2O (Bothner-By and Friedman 1952; Liu et al. 2014):

    $${\text{NH}}_{2} {\text{OH}} + {\text{HNO}}_{2} \to {\text{N}}_{2} {\text{O}} + 2{\text{H}}_{2} {\text{O}}{.}$$
    (22)

It is worth noting that when using chemical reagents (sodium azide in an acetic acid buffer or hydroxylamine hydrochloride) to reduce NO2− to N2O, it is necessary to add sodium arsenite solution to remove excess BrO− reagent before adding the next chemical reagent. Otherwise, the remaining BrO− will react with the next reducing agent, causing an interference (Liu et al. 2017).

3.3 Determination Nitrogen Isotopes in Ammonia

Passive samplers are commonly used for collecting dry N deposits because the samplers are inexpensive, easy to use, and do not require power. The passive samplers commonly used for NH3 sampling include those made by Ogawa and ALPHA (adapted low-cost passive high absorption)—these have been used many times to collect and measure NH3 (Felix et al. 2013). For passive NH3 sampling, polyamide fiber filters are typically impregnated with citric acid and then used to collect the samples. Compared with other acids, such as phosphoric acid or oxalic acid, the NH4+ concentration gap of citric acid is lower, and that makes it more suitable for high time-resolution measurements (Felix et al. 2012, 2013; Walters and Hastings 2018).

After a suitable sampling period, the NH3 on the filter from the passive sampler is extracted with ultrapure water. Aliquots of the extract can then be used to determine the nitrogen isotope composition of NH3 (δ15N-NH3) by the same methods as described above for NH4+. In addition, as there are no interferences from NO3− or NO2− in the passively collected NH3 sample, the NH3 nitrogen isotope composition can be determined by the denitrifying bacteria method (Felix et al. 2013).

4 Research Progress for the Determination of Stable Isotopes in Ammonia and Ammonium

The currently used NH3 collection methods have not yet been fully verified relative to the characterization of δ15N-NH3, but Walters and Hastings (2018) have discussed the suitability of an established collection device for that purpose. The device uses a honeycomb denuder (HCD) that is acid-coated (2% citric acid (w/v) + 1% glycerol (w/v) in a 80:20 methanol to water solution) and housed in ChemComb Speciation Cartridge (CCSC) (Koutrakis et al. 1993, 1988; Walters and Hastings 2018). The polytetrafluoroethylene (PTFE) coating inlet on the device can have a NH3 transmission efficiency > 97.3%, and a circular reservoir on the PM2.5 impactor is covered with PTFE to prevent particle bounce and to limit the absorption of NH3. In practice, two HCDs are used in series, and they are separated by a PTFE spacer: the first HCD is used to capture NH3 from the sample airstream, and the second is used as a control to check for possible NH3 breakthrough. The subsequent chemical treatment for N isotope determinations is carried out using the chemical conversion method with sodium azide in an acetic acid buffer (Zhang et al. 2007).

Walters et al. evaluated the reagent blanks, matrix effects, collection efficiency, selectivity, operative capacity, the stability of reaction products, field utility, and isotope bias for this device (Perrino et al. 1990), and they concluded that the method is suitable for the determination of δ15N-NH3. The acid-coated HCD collection device for δ15N-NH3 analysis and the subsequent chemical treatment lead to extremely high accuracy, with a precision of ± 1.6‰ (2σ; n = 75). The device has an operational capacity (collection efficiency > 95%) of 400 μg of collected NH3 at a concentration of ≤ 207 ppbv. In addition, this method is also applicable to studies of the natural atmosphere, and the resolution of δ15N-NH3 is ~ 1 h (Walters and Hastings 2018). This sampler represents the first laboratory-proven δ15N-NH3 device that has high time resolution and excellent δ15N-NH3 accuracy, and it may well become the sampler of choice for the air quality studies needed to improve our understanding of NH3 sources and sedimentation processes.

If a filter suitable for collecting PM2.5 is used, the HCD device could simultaneously measure δ15N-NH3 and δ15N-NH4+, making it useful for studying NHx. In addition, the combined denuder-filter collection technique may be extended to isotope analysis of other gaseous compounds such as sulfur dioxide (SO2), nitric acid (HNO3), and nitrogen dioxide (NO2) because, with a suitable HCD coating, the device may be used to collect different gases. For example, NH3 can be collected with the use of an acidic coating on the HCD, and acid gases such as HNO3 or SO2 can be collected by a carbonate coating on the HCD.

5 Conclusions

Due to the frequent haze events in China, there has been increasing interest and concern over the behavior of atmospheric NO3− and NH4+. Analysis of the stable N isotopes of atmospheric NO3− and NH4+ could provide a powerful tool to identify NOx or NH3 sources. Combined with some models, the contribution of different sources can be qualitatively and quantitatively assessed by isotope techniques. With this tool, it is possible to control and reduce the emission of nitrogen-containing pollutants from the emission sources. Therefore, it is necessary to review the techniques used for measurements of N isotopes in atmospheric aerosols. However, there also can be considerable overlap in the δ15N values for different emission sources. This suggests that there should be further exploration of N isotopes of major emission sources.