1 Introduction

Ammonia (NH3) is an alkaline nitrogen gas and prominent constituents of nitrogen cycle which plays a vital role in the neutralisation of acidic species and the formation of secondary aerosols in the atmosphere (Stockwell et al. 2000; Aneja et al. 2009; Saraswati et al. 2019a, b; Aneja et al. 2001; Huang et al. 2010, 2011; Seinfeld and Pandis 2006). The formation of secondary aerosols [(NH4)2SO4, NH4NO3 and NH4Cl] in the atmosphere influenced by reaction rate of NH3 which depends on the favourable meteorological condition (relative humidity and temperature) and level of acid gases (Ianniello et al. 2010, 2011; Meng et al. 2011; Saraswati et al. 2019b). These ammonium-containing aerosols constitute the major fractions of PM2.5 in the atmosphere which have major effect on human health (Matsumoto and Tanaka 1996; Sharma and Behra 2010; Updyke et al. 2012). Agricultural activities, livestock, transport and industrial activities majorly contribute NH3 to the atmosphere (Sutton et al. 2000; Li et al. 2006; Sharma et al. 2010a, b, 2011; 2016; Sutton et al. 2013; Xu and Penner 2012; Yang et al. 2011).

Fine fraction of particulate matter i.e. PM2.5 (diameter ≤ 2.5 µm) is considered as one of the major pollutants having a negative impact on atmospheric chemistry (Pant and Harrison 2012; Sharma et al. 2021). The anthropogenic activities contribute significantly to the mass concentration of total PM2.5 loading through gas-to-particle conversion (Huang et al. 2014). Secondary aerosols contribute to a major fraction of PM2.5 concentrations which is mainly formed from NH3 and its co-pollutants such that NOx and SOx (Sharma and Behra 2010; Saraswati 2019b; Singh et al. 2017). NH3 as a primary alkaline gas neutralises the acid gases (mainly HNO3 and H2SO4) and forms the secondary particulates (NH4NO3 and (NH4)2SO4) (Pinder 2007; Sharma et al. 2014b), which are the major fractions of airborne fine particles (Chow et al. 1994; Aneja et al. 2001; Huang et al. 2011).

In recent past, several studies on temporal and spatial changes of ambient NH3, NO, NO2, CO and SO2 have been carried on short-term basis as well as year-long basis at the urban and sub-urban locations of India (Khemani et al. 1987; Singh and Kulshrestha 2014; Kulshrestha et al. 1996; Parasar et al. 1996; Parmar et al. 2001; Sharma et al. 2010a, b; 2012a, 2012b, 2012c; 2014a, b, c, d; 2017; Saraswati et al. 2018; 2019a, b); however, Long-Term study on seasonal basis as well as gas-to-particle conversion is inadequate. In this paper, we report the annual and seasonal changes of ambient NH3, NO, NO2, SO2 and PM2.5 measured for the period of 2013–2018. We also emphasise the role of ambient NH3 and other trace gases (NO, NO2, SO2 and HNO3) in the formation of secondary aerosols in Delhi, India.

2 Methodology

2.1 Study Site

Ambient NH3, NO, NO2, and SO2 were monitored at CSIR-National Physical Laboratory (CSIR-NPL; 28° 38′ N, 77° 10′ E), New Delhi from January 2013 to December 2018 (Fig. 1). 24 h periodic sampling (2 samples/week) of PM2.5 was also performed during this period. Monitoring site represents as a typical urban location surrounded by nearby road traffic and ICAR-Indian Agricultural Research Institute (ICAR-IARI), New Delhi. Delhi experienced four distinct seasons i.e. winter (January–February), summer (March–May), monsoon (June–September) and post˗monsoon (October–December) as per meteorological classification. In Delhi region, winter months are chilly (temperature: ~ 2 °C) and noticeable intense fog and haze, whereas summers are generally very hot & dry (temperature: 47 °C) and experience repeated dust storms. A brief information about the study site is available in our previous paper (Sharma et al. 2021).

Fig. 1
figure 1

Map of the study site in Delhi (Source: Google maps)

Full size image

2.2 Monitoring of Trace Gases and PM2.5

Ground-based analyzers were used to continuous measurement of trace gases (NH3, NO, NO2 and SO2) at 10 m height from the surface level. In this study, NH3 analyzer was used (Model: AC3M&CNH3, M/s. Environnment SA, France) to measure the mixing ratios of NH3, NO and NO2 (working on chemiluminescence method with accuracy ± 1.0 ppb). Ambient SO2 was used to measure SO2 analyzer (Model: APSA 360A, M/s. Horiba Ltd., Japan) operating on ultraviolet fluorescence method (accuracy ± 0.5 ppb). Zero, span checks as well as periodic calibrations of these analyzers were performed using Zero Air Generator (Model: PAG-003, M/s. ECO Physics AG, Switzerland, accuracy ± 0.01 ppb) and NIST-certified respective reference gases (Saraswati et al. 2019b). The detailed principle of operations, calibration procedures and level of certified standard gases used are discussed in Sharma et al. (2014b). HNO3 samples were collected using the standard impinger containing de-ionised water (25 ml) from November 2014 to December 2015 and analysed using Ion Chromatograph (DIONEX-ICS-3000, USA). Prebaked (at 550 °C for 5 h) and desiccated quartz microfiber filters (diameter: 47 mm) were used to collect PM2.5 samples using fine particle sampler (APM 550, Make: M/s. Envirotech, India) for 24 h [at a flow rate of 1 m3 h−1 (accuracy ± 2%)] from January 2013 to December 2018 (Saraswati et al. 2019b).

2.3 Chemical Analysis

To estimate the Water-Soluble Inorganic Components (WSICs) of PM2.5, the collected filters were extracted in ultra-pure water for 90 min and extracted through 0.22 µm nylon filter. Cations (Li+, Na+, NH4+, K+, Ca2+, and Mg2+) and anions (F−, Cl−, NO3− and SO42−) of PM2.5 concentrations were determined using Ion Chromatograph (DIONEX-ICS-3000, USA) with suppressed conductivity. The field blank filters were also analysed for blank correction. The detailed procedure of cations and anions analysis, calibration of ions, standards used and repeatability errors are available in our previous publication (Sharma et al. 2014b).

2.4 Meteorological Data

The ambient temperature (°C), relative humidity (RH; %), wind speed (ms−1) and wind directions (degree) were also monitored using sensors placed on meteorological tower (5 stages tower of 30 m height). In this study, we used the meteorological data (temperature, RH, wind speed and wind direction) available at 10 m height (AGL) and summarised in Table 1 and Table S1 (see the supplementary information).

Table 1 Annual mixing ratios of trace gases and meteorological parameters
Full size table

2.5 Estimation of NH4 + availability index (J) and conversion

For the observational site, NH4+ availability index (J) was estimated to explore the availability of NH4+ for neutralisation of acidic components of acid gases (H2SO4, HNO3 and HCl) during the study period (Adam et al. 1999; Chu 2004) and is expressed as:

$$J = \frac{{\left[ {NH_{4}^{ + } } \right]}}{{2 \times \left[ {SO_{4}^{2 - } } \right] + \left[ {NO_{3}^{ - } } \right] + \left[ {Cl^{ - } } \right]}}\,\, \times \,\,100\%$$
(1)

When J < 100%, it means it is an NH4+ deficit condition, which indicates that SO42–, NO3– and Cl− are acidic. When J = 100%, the particulate is neutral, indicating precise neutralisation of SO42–, NO3– and Cl−. When J > 100%, there is a sufficient NH4+ to fully neutralise acidic SO42–, NO3– and Cl−.

In the present study, using the seasonal average of gaseous NH3 and particulate ammonia of PM2.5, the % fraction of N–NH4+ and N–NH3 was computed for different seasons using the following equation:

$$\% N - NH_{4}^{ + } = \frac{{\left[ {N - NH_{{4\left( {{\text{Aerosol}}} \right)}}^{ + } } \right]}}{{\left[ {N - NH_{{4\left( {{\text{Aerosol}}} \right)}}^{ + } } \right] + \left[ {N - NH_{{3 \left( {{\text{Ammonia}}} \right)}} } \right]}} \times 100\%$$
(2)

3 Results and Discussion

3.1 Mixing Ratios of Trace Gases and WSIC of PM2.5

The annual average levels of trace gases (NH3, NO, NO2 and SO2), WSICs and meteorology at the observational site of Delhi from January 2013 to 2018 are depicted in Table 1. During the entire study period (2013–2018), the average levels of NH3, NO, NO2, SO2 and HNO3 were 19.1 ± 3.8, 20.8 ± 4.3, 17.9 ± 4.2, 2.45 ± 0.47, 1.11 ± 0.35 ppb, respectively, whereas the levels of NH4+, SO42−, NO3− and Cl− of PM2.5 were 9.1 ± 3.5, 12.3 ± 4.1, 10.8 ± 4.8 and 9.3 ± 3.2 µg m−3, respectively. The highest mixing ratio of ambient NH3 was recorded in 2014 (24.4 ± 4.5 ppb) and the lowest level of NH3 in 2016 (15.9 ± 9.1 ppb). The inter-annual variability in ambient NH3, NO, NO2, CO and SO2 levels were discussed in detail in our previous publication and reference therein (Sharma et al. 2017). The annual concentrations of PM2.5 estimated as 135 ± 45 µg m−3 with a range of 35–451 µg m−3. The annual concentration of PM2.5 at the sampling site of Delhi exceeded more than 3 times of National Ambient Air Quality Standard (NAAQS; annual: 40 µg m−3) of India and more than 25 times of World Health Organisation (WHO) guideline (5 µg m−3).

Seasonal mixing ratios of NH3, other trace gases (NO, NO2 and SO2) and concentrations of WSICs of PM2.5 are depicted in Table 2, whereas the average seasonal variation in meteorology of observational site is summarised in Table S1 (in supplementary information). The monthly averages (pooled average from 2013 to 2018) of trace gases were also depicted in Fig. 2, whereas monthly average time series of these trace gases were also reported in Fig. S1 (in supplementary information). The ambient NH3 indicated significant seasonal variation with highest mixing ratio during the post-monsoon season (22.2 ± 3.9 ppb) followed by the winter (20.9 ± 4.1 ppb), summer (19.4 ± 4.1 ppb) and monsoon (14.0 ± 2.5 ppb) seasons. The average levels of NO, NO2 and SO2 were also recorded highest during the post-monsoon and lowest during  the monsoon season except SO2 (Table 2). The average concentrations of NH4+ were recorded as 17.5 ± 2.8 µg m−3, 9.3 ± 4.4 µg m−3, 5.8±3.5 µg m−3 and 3.9 ± 1.2 µg m-3 during the winter, post-monsoon, summer and monsoon seasons, respectively. The higher concentration of NH4+ (17.5 ± 2.8 µg m−3) during the winter season at the observational site of Delhi may be due to high RH (62.4 ± 12.6 %), low temperature (16.7 ± 3.6 °C) and higher NH3 (20.9 ± 4.1 ppb) mixing ratio influenced the NH4+ aerosols formation (Khoder 2002). Similar seasonal concentrations of SO42− and NO3− were also recorded with highest concentration during the winter season (Table 2). The higher source strength of mixing ratios of precursor gases, such as NH3, NOx and SO2, may lead to high concentrations of NH4+, SO42− and NO3− during the winter season. The higher RH during  the winter season also favours dissolution of significant fraction of NH3 which increases the NH4+ formation (Ianniello et al. 2010).

Table 2 Seasonal variation in trace gases (in ppb) and WSIC of PM2.5 (in µg m−3) in Delhi during 2013–2018
Full size table
Fig. 2
figure 2

Monthly average (pooled average of 2013–2018) mixing ratios of NH3, NO, NO2 and SO2 in Delhi, India

Full size image

In this study, the higher concentration of HNO3 was estimated in the summer season followed by the winter, post-monsoon and monsoon seasons (Table 3). The more availability of HNO3 in the summer at the study might be due to increase photochemical activity and higher level of OH radical (Hoek et al. 1996; Wu et al. 2009) which increased the availability of HNO3 from gaseous precursors gases (NO2 + OH = HNO3) (Behera and Sharma 2010; Liu et al. 2015). Sharma et al. (2007) also reported the 3 times higher HNO3 level in summer than the winter at IIT Kanpur over the IGP region of India.

Table 3 The measured concentration product (Km) and theoretical equilibrium constant (Kp) of NH4NO3 formation during winter, summer and monsoon seasons during the measurement period at the study site
Full size table

Mass concentration of PM2.5 was recorded higher in the winter (190 ± 82 µg m−3) followed by the post-monsoon (171 ± 72 µg m−3), summer (92 ± 30 µg m−3) and monsoon (86 ± 33 µg m−3) seasons (Table 2). One of the reasons of the high level of PM2.5 during winter could be the higher formation of secondary particles like NH4+ (17.5 ± 2.8 µg m−3), SO42− (19.6 ± 6.9 µg m−3), NO3− (22.7 ± 9.5 µg m−3) and Cl− (15.6 ± 8.9 µg m−3) during  the winter season (Table 2). During winter, the higher level of Cl− was recorded than other seasons (post-monsoon, summer and monsoon) due to enhanced fossil fuel combustion (Wang et al. 2005), whereas K+ ion was higher due to biomass burning in Delhi and their surroundings (Sharma et al. 2014b). The higher concentrations of all the pollutants in the colder season may be due to the lower mixing height [100–250 m at sampling site as reported by Kumar et al. (2021)] by surface temperature inversion (Xu et al. 2016; Wang et al. 2014) and weak winds for lower dispersion of the pollutants. During all the seasons, mixing ratio of ambient NH3 is negatively correlated with ambient temperature (winter: r2 = − 0.49, summer: r2 = − 0.76, monsoon: r2 = − 0.69, post-monsoon: r2 = − 0.75; p < 0.05), whereas positively correlated with the RH (winter: r2 = 0.54, summer: r2 = 0.74, monsoon: r2 = 0.75, post-monsoon: r2 = 0.59; p < 0.05). Similar correlations were also reported between these gases and meteorological parameters at other urban site (Makovic et al. 2008; Sharma et al. 2010a, b, c). The wind is one of the most important parameters for the transfer and dispersion of pollutants. The increase in wind speed indicates the increase in transport of pollutants which may result in lower pollutant concentration. In the present study, ambient NH3 was negatively correlated with wind speed during all the season (winter: r2 = − 0.67, summer: r2 = − 0.47, monsoon: r2 = − 0.75, post-monsoon: r2 = − 0.76; p < 0.05). The highest NH3 mixing ratio was observed during the lower  wind speed (1–2 m s−1) from downwind direction indicates the possible nearby sources and lowest NH3 at high wind speed (5–6 m s−1). The higher NH3 mixing ratio was observed during the winter season may also be due to the lower wind speed which associated with local sources from nearby agricultural field.

3.2 Ionic Composition and Ionic Balance

NH3 reacting with acid gases (H2SO4, HNO3 and HCl) forms the NH4SO4, NH4NO3 and NH4Cl compounds which are referred as secondary inorganic aerosols (SIA). The secondary inorganic particulate components (NH4+, SO42−, NO3− and Cl−) were dominant in PM2.5 at the sampling location of Delhi. The secondary inorganic aerosol components i.e. NH4+, SO42−, NO3− and Cl− manifested the highest levels in the PM2.5 with an average level of 9.1 ± 3.5, 12.3 ± 4.1, 10.8 ± 4.8 and 9.3 ± 3.2 µg m−3 during the study period (2013–2018). The NH4+, SO42−, NO3− and Cl− were the major portions of the total soluble ions (Table 2). In winter, nitrates availability was significant due to possible reduction in SO2 oxidation rates in response to lower level of OH radical (Walker et al. 2004). A relationship of particulate NH4+ with SO42−, NO3− and Cl− during all the seasons supports the hypothesis (Fig. 3).

Fig. 3
figure 3

Scatter plots between NH4+ vs. SO42-, NH4+ vs. NO3−, and NH4+ vs. Cl− of PM2.5 in Delhi during winter, summer, monsoon and post-monsoon seasons

Full size image

The molar ratios of NH4+ with SO42−, NO3− and Cl− of PM2.5 were also computed and summarised in Table 4, whereas charge balances are depicted in Fig. 4. It is to be noted that the molar ratio of NH4+/SO42− < 2 indicates the NH4+ poor condition and > 2 indicates the NH4+ rich condition at the sampling site. The highest average molar ratio of NH4+ to the SO42− during winter (4.86) followed by post-monsoon (4.38), summer (3.61) and monsoon (2.1) seasons indicated the complete neutralisation of H2SO4, abundance of (NH4)2SO4 and NH3-rich condition during the winter season (Saraswati et al. 2019b). Since NH3 is the only alkaline gas in the atmosphere with adequate level to neutralise a significant portion of SO42−, NO3− and Cl−, therefore, the aerosol electro-neutrality relationship (NH4+ availability index: J) between NH4+ and SO42−, NO3− and Cl− ions can be computed (Chu 2004; Behra and Sharma 2010). During the winter (125.8%), post-monsoon (125.6%) and summer (102.2%) seasons, the average value of J was > 100% at the sampling site of Delhi. So there was enough NH4+ present in PM2.5 samples to neutralise the SO42–, NO3– and Cl− (Behra and Sharma 2010; Adam et al. 1999).

Table 4 Model ratios NH4+ with other ionic species
Full size table
Fig. 4
figure 4

Charge balance: a between SO42− vs. NH4+, b between SO42− + NO3− vs. NH4+ and c between SO42− + NO3− + Cl− vs. NH4+ during winter, summer, monsoon and post-monsoon in Delhi

Full size image

In the present case, the particulate NH3 (in PM2.5) had been lower than the gaseous NH3 during all the seasons except the winter (54%) season (summer: 29%, monsoon: 28% and post-monsoon: 37%). Similar results were also reported by Singh and Kulshrestha (2012) at the other urban location of Delhi. The % fraction of (N-NH4+) was higher in the winter (54%) than all the other seasons. The higher N–NH4+ fraction during the winter may be due to favourable meteorological condition (lower temperature and higher RH) and abundance of NH3 at sampling site, which results in faster NH3 to NH4+ conversion (Saraswati et al. 2019b).

3.3 Gas to Particle Conversion

In the atmosphere, the acid gases (HNO3 and H2SO4) formed by the reaction of NOx and SOx with hydroxyl radical (OH) and the particulate NH4+ were formed by the reaction of gaseous NH3 with HNO3 and H2SO4 acids gases (Heeb et al. 2008). In the present study, the role of atmospheric NH3 in the formation of NH4NO3 and NH4SO4 was examined by estimating the significant positive correlation of NH3 with NOx and SO2 during  the winter (r2 = 0.84, r2 = 0.52; p < 0.05), summer (r2 = 0.82, r2 = 0.54; p < 0.05), monsoon (r2 = 0.52, r2 = 0.47; p < 0.05) and post-monsoon (r2 = 0.79, r2 = 0.58; p < 0.05) seasons, respectively (Table S2 a–d, see the supplementary information). The positive correlation of NH3 with NH4+ also indicates the role of NH3 in the transformation of NH4+ during all the seasons (winter: r2 = 0.77; summer: r2 = 0.49; monsoon: r2 = 0.52; post-monsoon: r2 = 0.51; p < 0.05). The gaseous ammonia was converted into particulate ammonia (of PM2.5) and estimated as 54, 29, 28 and 37% during the winter, summer, monsoon and post-monsoon, respectively which further neutralised the SO42–, NO3– and Cl− particulates. The average NH3/NH4+ ratios varied from 1.19 to 3.58 with an average level of 2.62 during entire study period. The average NH3/NH4+ ratio was computed as 1.19, 3.34, 3.58 and 2.38 during the winter, summer, monsoon, and post-monsoon, respectively. It was observed that the acidic gases (HNO3 and H2SO4) were completely neutralised by the NH3 gas during winter (as NH3/NH4+ was close to 1). The higher ratio of NH3/NH4+ during the  summer, monsoon and post-monsoon suggested that the NH3 gas was not neutralised completely (Meng et al. 2011) and NH3 remained predominantly in the gas phase rather than aerosols phase (gaseous NH3 to total NHx (NHx = NH3 + NH4+  = 0.68) (Gong et al. 2013). The higher seasonal NH4+/NHx ratios during all the seasons reflect the higher formation of NH4+ due to low temperature and higher RH (Saraswati et al. 2019b). Behera et al. (2013) also demonstrated the formation of particulate NH4+ from NH3 at urban site of Kanpur, India.

The sulphur oxidation ratio [SOR = SO42−/(SO42− + SO2)] and nitrogen oxidation ratio [NOR = (PNO3− + GNO3)/(NO2 + PNO3− + GNO3−)] are used as indicators of the secondary transformation process and sources of SO42− and NO3−, respectively (Khodar 2002; Baek and Aneja 2004; Behera and Sharma 2010). The higher SOR and NOR values were recorded in the winter (0.72, 0.43) followed by the post-monsoon (0.61, 0.27), summer (0.59, 0.22) and monsoon (0.58, 0.16) seasons at the sampling site (Table 5). Behera and Sharma (2010) had also obtained the higher level of SOR than NOR during the winter and summer seasons at Kanpur, India which demonstrates the low formation of NO3− as compare to SO42−. Saxena et al. (2017) also reported the higher level of SOR and NOR during winter in Delhi, whereas Sharma et al. (2007) showed the high NOR due to enhanced NO3− formation during winter influenced by high humidity at Kanpur.

Table 5 Values of SOR and NOR during different seasons in Delhi
Full size table

In the above section, the significant positive correlation of NH4+ with SO42− and NO3− was observed which indicates the formation of NH4SO4 and NH4NO3 and their contribution to PM2.5 (Table S2 a–d; in the supplementary information). We had also examined the reversible reaction of NH3 and HNO3 which results in the formation of particulate NH4NO3 as:

$${\text{NH}}_{{3}} \left( {\text{g}} \right) + {\text{HNO}}_{{3}} \left( {\text{g}} \right) \leftrightarrow {\text{NH}}_{{4}} {\text{NO}}_{{3}} \,\left( {\text{S or aq}} \right)$$
(3)

The equilibrium of NH4NO3 between gas and particulate phase depends on ambient temperature and RH (Wei et al. 2015). In this case, the product of measured concentration (Km = [NH3] * [HNO3]) was computed and examined with the theoretically calculated equilibrium constant (Kp) at the same meteorological condition (Stelson and Seinfeld 1982; Saraswati et al. 2019b) and depicted in Table 3. The value of thermodynamic equilibrium constants calculated for the  summer and the monsoon seasons indicated that NH4NO3 is not expected to be formed at the observational site (may be due to at most of the days as the product of NH3 and HNO3 were below the thermodynamically predicted dissociation constant) (Table 3). The conditions in the winter and the post-monsoon were more favourable for the formation of NH4NO3 as compared to summer and monsoon seasons (Behera and Sharma 2010).

4 Conclusion

In this paper, the Long-Term average seasonal mixing ratios of ambient trace gases (NH3, NO, NO2, and SO2) and WSICs of PM2.5 (NH4+, SO42−, NO3- and Cl− etc.) were estimated at the observational site of Delhi, India (January 2013–December 2018) to examine the role of ambient NH3 in the formation of secondary inorganic aerosols at the study site. The average levels of all trace gases (NH3, NO, NO2 and SO2) were observed higher during the post-monsoon season, whereas the mass concentrations of WSICs of PM2.5 were higher in the winter season. The correlation matrix of trace gases demonstrated that the ambient NH3 neutralised all acid gases (NO, NO2 and SO2) at Delhi during the study period. Ion balance and molar equivalent ratios analysis of PM2.5 also indicated that the abundance of particulate NH4+ at the study site to neutralise the SO42−, NO3−, Cl− particles during all seasons, whereas the formation of NH4NO3 was higher during the winter due to favourable meteorological condition and forward reaction of NH3 and HNO3.