Introduction

Volatile organic compounds (VOCs) are the major precursors of ozone (O3, secondary pollutant) and play an important role in tropospheric chemistry. Emissions of VOCs have natural (vegetation and seawater) as well as anthropogenic sources in the troposphere. The global emission rate is estimated at ~1 × 1014 gC yr−1 for anthropogenically derived compounds (stationary and mobile sources) (Piccot et al. 1992) and ~8 × 1014 gC yr−1 for biogenic compounds (Guenther et al. 1995; Zimmerman et al. 1978). In Hyderabad city, the air pollution load is also mainly contributed by anthropogenic emissions due to rapid urbanization and increased economic activity. The present experimental work was carried out at Tata Institute of Fundamental Research-National Balloon Facility (TIFR-NBF), Hyderabad, to analyze the relative variability in VOCs concentrations. Ambient air samples were collected intermittently at TIFR and from different surrounding sites viz., ECIL X Roads (2 km), Tarnaka (10 km), and Habsiguda (12 km) which are away from the experimental site. The emission sources at these sites are majorly contributed by automobile traffic (both light and heavy motors) and liquefied petroleum gas (LPG) leakage from the nearby gas filling stations. The reports by Census India 2011 (http://www.census2011.co.in/census/district/122-hyderabad.html) recorded the population density of 18,172 people per square kilometer for Hyderabad. This city comprises of many industrial development areas and the main experimental site is surrounded by many industries such as Electronics Corporation of India Limited (ECIL), Hindustan Cables Limited (HCL), Nuclear Fuel Complex (NFC), petroleum storage containers, bottling units of Hindustan Petroleum Corporation Limited (HPCL), and Bharat Petroleum Corporation Limited (BPCL). The records of central pollution control board (CPCB) showed that there were over 1.1 million vehicles on road in this city in the year 2006 (http://cpcb.nic.in/). Guttikunda and Kopakka (2014) reported the sector-specific emissions for 2010–2011 for this city, accounted for 113,400 t of VOCs emissions. The predominant anthropogenic sources of VOCs are vehicular and industrial emissions such as fossil fuel combustion, biomass burning, LPG and natural gas (NG) leakages, fuel evaporation, petroleum distillation, oil refineries, and solvent usage (Barletta et al. 2005; Chen et al. 2001; Cheng et al. 1997; Duan et al. 2008; Na et al. 2001; Seila et al. 2001; Tang et al. 2007; Watson et al. 2001). In polluted urban areas, contribution of alkanes (45 %) in ambient air is higher than that of alkenes (10 %), and aromatic hydrocarbons (20 %) towards VOCs (Solomon 1994). Biogenic VOCs like isoprene also significantly contribute in the urban and regional environments. Anthropogenic and biogenic VOCs are important precursors for secondary organic aerosol (SOA) formation, where organic component of particulate matter transfers to the aerosol phase from the gas phase as products of gas-phase oxidation of parent organic species in polluted regions (Kanakidou et al. 2005).

A large number of complex photochemical reaction take place by wide range of VOCs with different reaction mechanisms in the troposphere at urban and rural areas (Carter 1994). The photochemical oxidation of VOCs initiated by rapid reaction with hydroxyl radicals (OH˙) involves various chemical chain reactions and several gaseous intermediates such as alkyl peroxy radicals (RO2˙), peroxy acetyl nitrate (PAN), carbonyls and organic nitrate compounds (RONO2). Photochemical processes involving VOCs in the presence of sufficient amount of nitrogen oxides (NO x ) leads to O3 formation (Caselli et al. 2010; Duan et al. 2008; Seinfeld 1989; Sillman 1999; Steinfeld 1998; Han et al. 2011; Swamy et al. 2012; Yerramsetti et al. 2013b). Further, the reaction of OH˙ with intermediate oxidation products (acetone, aldehydes, hydroperoxides etc.,) enhances the lifetime of greenhouse gases (GHGs) (Offenberg et al. 2011; Poisson et al. 2000). The reactivity of each VOC determines its respective impact over regional atmospheric chemistry. Peroxy radicals (HO2˙) can be formed as a result of photolysis processes. In brief, O3 is formed through oxidation of hydrocarbons by forming peroxy radicals according to the following chain reactions.

$$ \begin{array}{l}{\mathrm{CH}}_3-{\mathrm{CH}}_3+{\mathrm{OH}}^{\cdotp}\to {\mathrm{CH}}_3-{{\mathrm{CH}}_2}^{\cdotp }+{\mathrm{H}}_2\mathrm{O}\hfill \\ {}\left\{\mathrm{one}\ \mathrm{example}\ \mathrm{f}\mathrm{o}\mathrm{r}\ \mathrm{alkyl}\ \mathrm{r}\mathrm{adical}\left({\mathrm{R}}^{\cdotp}\right)\right\}\hfill \end{array} $$
(1)
$$ {\mathrm{CH}}_3-{{\mathrm{CH}}_2}^{\cdotp }+{\mathrm{O}}_2\to {\mathrm{CH}}_3-{\mathrm{CH}}_2{{\mathrm{O}}_2}^{\cdotp } $$
(2)

Alkyl peroxy radicals and hydro peroxy radical can oxidize NO to NO2 to form hydroxyl radical.

$$ {\mathrm{CH}}_3-{\mathrm{CH}}_2{{\mathrm{O}}_2}^{\cdotp }+\mathrm{NO}\to {\mathrm{CH}}_3{\mathrm{CH}}_2{\mathrm{O}}^{\cdotp }+{\mathrm{NO}}_2 $$
(3)
$$ {\mathrm{CH}}_3{\mathrm{CH}}_2{\mathrm{O}}^{\cdotp }+{\mathrm{O}}_2\to {\mathrm{CH}}_3\mathrm{C}\mathrm{H}\mathrm{O}+{{\mathrm{HO}}_2}^{\cdotp } $$
(4)
$$ {{\mathrm{HO}}_2}^{\cdotp }+\mathrm{NO}\to {\mathrm{OH}}^{\cdotp }+{\mathrm{NO}}_2 $$
(5)

In the troposphere, NO2 is formed in several processes provides the primary source of oxygen atoms in presence of sunlight which is required for O3 formation.

$$ {\mathrm{NO}}_2+\mathrm{h}\upnu\ \left(<420\;\mathrm{nm}\right)\to \mathrm{NO}+\mathrm{O}\left({}^3\mathrm{P}\right) $$
(6)

The oxygen atom combines with an oxygen molecule to produce O3

$$ \mathrm{O}\left({}^3\mathrm{P}\right)+{\mathrm{O}}_2+\mathrm{M}\to {\mathrm{O}}_3+\mathrm{M} $$
(7)

The O3 formed in troposphere is one among the GHGs that shows adverse effects on human health and plant growth when exceeded above the ambient air quality standards (Burnett et al. 1994; Choi and Hwang 2011; Lai et al. 2009; Mudliar et al. 2010; Shao et al. 2009). In India, the National Ambient Air Quality Standards (NAAQ) has proposed 50 ppbv of O3 in the surface level (http://cpcb.nic.in/National_Ambient_Air_Quality_Standards.php). Hyderabad is on the most polluted city in south Indian region, which has many sources of hydrocarbons. However, there were no monitoring studies on concentration and speciation of VOCs in this tropical urban site. Therefore, an investigation was carried out to determine the factors influencing the temporal variations of VOCs and emission sources and empirically analyze the contribution of hydrocarbons towards O3 formation at this site. Estimation of the reactivity and contribution of individual hydrocarbons to photochemical O3 formation was performed by using reactivity scale methods such as the propylene-equivalent concentration (propy-equiv.) and the ozone formation potential (OFP). Maximum incremental reactivity (MIR), which is a good indicator for comparing OFP of individual hydrocarbons, was used for the study.

Experimental methodology

Analysis of O3 and NO x

Continuous monitoring of O3 and NO x was carried out using online gas analyzers which are installed at Tata Institute of Fundamental Research-National Balloon Facility (TIFR-NBF), Hyderabad, Telangana, India. The O3 analyzer (49i; Thermo Scientific, USA) operates on the UV light absorption principle at a wavelength of 254 nm and NO x analyzer (42i; Thermo Scientific, USA) works on chemiluminescence principle where light intensity is linearly proportional to the NO concentration. Accuracy of the analyzers was maintained by calibrating every fortnight with NIST (US National Institute of Standards and Technology) traceable standard gas by using multi-gas calibrator (Model 146i, Thermo Scientific, USA). Detailed methodology of trace gas measurements was previously reported (Swamy et al. 2012; Venkanna et al. 2014).

Analysis of VOCs

Ambient air samples were collected at the experimental TIFR site and sub-sites (one rural site and two traffic junctions, Habsiguda and Tarnaka, in Hyderabad) using a 500-mL glass air sampling unit. Measurements of ethane, propane, i-butane, n-butane, n-pentane, i-pentane, ethene, propene, and acetylene were carried out using the pre-concentration set-up connected to the gas chromatography (Shimadzu GC-17A, Japan). The collected air samples were transferred into an evacuated stainless steel canister of 250 ml. The canister was connected to an 8-valve port (V1), U-type column (25 cm long with 6.3 mm outer diameter) filled with glass beads of 60/80 mesh was connected between one port of V1 with another 6-valve port (V2). The air from the canister was loaded into the U column and pre-concentrated for a pre-determined time under cryogenic conditions by using liquid nitrogen. After pre-concentration, the column was kept under isolated condition and heated with warm water; later, the desorbed volatile gases were injected into GC. Sample loading, isolation, and injection were carried out by operating V1 and V2 valves. VOCs were analyzed using GC equipped with flame ionization detector (FID) and PLOT capillary column (Supelco) of 30 M long, 0.53 mm i.d. with Al2O3/KCl as stationary phase. High purity nitrogen (N2) was used as carrier gas. The column oven temperature was ramped from 40 to 180 °C. Initially, 40 °C was maintained for the first 5 min; afterwards, the temperature was raised at a rate of 5 °C/min until it reached 120 °C and kept at this temperature for 5 min. In the final step, the temperature was increased to 180 °C at a rate of 25 °C/min, and then maintained at 180 °C till the end of the analysis. Pure standard of the individual gases were injected to identify their retention times in chromatogram. Besides, for quantification of each peak, the calibration was done using two different calibration gas mixtures (NIST traceable) and found to be more reliable. GC calibration was done before analysis. The analytical performance was evaluated based on linearity and accuracy parameters. The minimum detection limits for all the nine VOCs were in the range of 50–150 pptv. The overall uncertainties were in the range of 10–20 % of detection limits for all the gases. A calibration graph of area versus concentration was drawn for the quantification of each air sample (Swamy et al. 2012).

Results and discussion

Diurnal variations of hydrocarbons

The annual mean diel variations of C2–C5 VOCs during different time intervals were evaluated for the TIFR site for the year 2013 and are graphically represented in Fig. 1. Vertical bars in the figure show the standard deviation (1 s) from the mean. Diel patterns of hydrocarbons showed morning and evening peak concentrations and low concentrations during afternoon hours (1200 to 1600 h). The maxima at morning/night time and minima at afternoon is a distinct feature of an urban polluted site. The VOCs concentrations during morning and evening are attributed to emissions from heavy vehicular traffic and weak vertical diffusion in the shallow boundary layer (Liu et al. 2015; Toro et al. 2014a, b). Low wind speed and high relative humidity during night time causes the weak diffusion of gases. The lower VOCs levels during noon time are due to the increased mixing boundary layer height and photochemical oxidation reaction with OH· radicals in presence of high solar radiation and temperature (Liakakou et al. 2009; Tang et al. 2007).

Fig. 1
figure 1

Annual mean diel variations of hydrocarbons at TIFR in the year 2013

Full size image

Ethylene and propene showed larger amplitude compared to other hydrocarbons almost throughout the year. Afternoon concentrations of ethylene and propylene were 51 and 54 % lower than the evening peak hour levels. These variations might be due to high reactivity of alkenes (ethene and propene) with hydroxyl radicals (OH·) predominantly in presence of high temperature and solar radiation compared to other hydrocarbons (Atkinson 1997; Hagerman et al. 1997). With increasing VOC reactivity, the diel amplitude (percentage of the nighttime maximum to the daytime minimum concentration) also increases. However, diel amplitude was less in case of n-pentane (18 %) compared to other hydrocarbons due to low reactivity towards photochemical oxidation as well as contribution by the emission of pentanes from gasoline and solvent evaporation at high temperatures at noon time (Barletta et al. 2005; Tan et al. 2012).

The average concentrations of nine hydrocarbons are placed in the order of propylene < n-butane < i-butane < ethylene < ethane < n-pentane < propane < i-pentane < acetylene. Among them, i-pentane, propane, and acetylene were the most abundant hydrocarbon with the average mixing ratio of 5.3 ± 2.1, 4.3 ± 0.9, and 5.5 ± 1.3 ppbv, respectively. LPG is used as an energy source in cooking, and leakages from LPG fuel-based vehicles are major sources for propane in Hyderabad (http://cpcb.nic.in/). Propylene and ethylene were observed to be relatively low due to less life time of these unsaturated HCs than alkanes. High mixing ratios of acetylene might be associated with its high life time among unsaturated HCs.

Interspecies correlations of hydrocarbons

The interspecies correlation study of VOCs is useful in comparing important specific sources of VOCs. The correlation plots between the VOC pairs for hourly mean of different seasons are shown in Fig. 2, at TIFR site. The relations between propylene vs. ethene and acetylene showed high coefficients of determination about 0.94 and 0.81, respectively. These strong correlations suggest, they have in common emission sources such as fossil fuel combustion of vehicles. The high correlation of propylene with ethylene (r 2 = 0.94) can be explained by their similar reactivities (KOH (propylene) = 26.3 × 1012 cm3 molecule−1 s−1 and KOH (ethene) = 8.52 × 1012 cm3 molecule−1 s−1) compared to acetylene (KOH = 0.9 × 1012 cm3 molecule−1 s−1) with atmospheric hydroxyl radicals (Atkinson 1986, 1997; Atkinson and Arey 2003). Similarly, Fig. 2a shows strong correlations between propane vs. other alkanes. The high correlations between propane vs. ethane (r 2 = 0.82), i-butane (r 2 = 0.84), n-butane (r 2 = 0.85), i-pentane (r 2 = 0.72), and n-pentane (r 2 = 0.71) can be explained by their common source of emissions and their similar rates (rate constant for the reactions of OH with alkanes presented in Table 1) of atmospheric degradations. Since LPG leakage is the largest common source of atmospheric propane and butanes in the urban areas (Blake and Rowland 1995), propane showed high correlation with butane (r 2 = 0.84 and 0.85). As LPG leaking is not a major source of pentane, it showed relatively weak correlation with propane (0.72 and 0.71) compared to butanes. Further, the emission ratios derived from the TIFR site data are compared with the emission ratios of vehicle exhaust (Harley et al. 1992) presented in Fig. 3. The emission ratios for C2–C5 alkanes are higher at TIFR site than the vehicle exhaust, confirming the emission of alkanes from additional sources, possibly from the LPG leakage and gasoline evaporations (De Gouw et al. 2005). Further, weaker correlation was observed between alkenes (propylene) vs. alkanes such as propane, i-butane, n-butane, i-pentane, and n-pentane with coefficients of determination about 0.68, 0.60, 0.61, 0.56, and 0.57, respectively. Weak correlation of alkenes with alkanes confirmed their different emission sources and reactivities of these species at this site.

Fig. 2
figure 2

Interspecies correlation study of VOCs between a alkanes vs. alkanes, b alkanes vs. alkenes, and c alkenes vs. alkenes

Full size image
Table 1 Annual mean propy-equiv. and OFP of nine hydrocarbons (TIFR site)
Full size table
Fig. 3
figure 3

Comparison of the VOCs emission ratios at TIFR site and vehicular exhaust (Harley et al. 1992)

Full size image

Seasonal variations of hydrocarbons

Based on the climatology of this area, the data was classified into three major different seasons viz., summer (March to June), monsoon (July to October), and winter (November to February). The seasonal variations of hydrocarbons at TIFR site are shown in Fig. 4. VOCs showed distinct seasonal variations. It is clear that mean concentrations of these hydrocarbons were very high in winter followed by summer and monsoon. Summer showed low mean concentration of all hydrocarbons, slightly increased in monsoon with maximum in winter season. The winter concentrations were 181 and 155 % higher than summer and monsoon, respectively. The seasonal variations of air pollutants in Hyderabad are mainly influenced by meteorological conditions. Monsoonal rain brings in clean marine air from South-West direction (Fig. 5), and in summer, increased boundary layer height play a significant role in the variations of O3 precursors. The low levels of VOC concentration during summer seasons were attributed to the increased OH radical chemistry with hydrocarbons in presence of high temperature and radiation resulting the formation of O3 and other secondary free radicals (Swamy et al. 2012). The low levels in monsoon season compared to winter were due to washout of pollutants. The arrival of continental air mass loaded with pollutants in winter from North and North-East part of India caused high levels of VOCs in winter (Venkanna et al. 2014). The varying lifetimes of hydrocarbon and atmospheric transport of air masses from different origins could contribute to the seasonal variations in VOC levels (Liakakou et al. 2009).

Fig. 4
figure 4

Variation in hydrocarbons during three seasons in the year 2013 at TIFR site

Full size image
Fig. 5
figure 5

Variations of wind speed and directions during different seasons for the year 2013

Full size image

Comparison of VOCs at different sites

Comparison study on the variation of concentration levels of hydrocarbons at different sites was performed by collecting samples at the respective sites viz., background rural site, traffic junctions in urban area, and TIFR experimental site. Average values of three traffic junctions were considered to represent hydrocarbon pollutant levels in traffic junctions at Hyderabad as VOCs emission ratios at all three sites were observed to be very high compared to TIFR site throughout the year. The annual mean concentration levels (± standard deviation) of nine VOCs concentrations at the different sites are represented in Fig. 6. Annual mean mixing ratios of different VOCs at TIFR site were higher than that of rural background due to heavy industrial emissions and contribution from local transport of vehicular emissions. These VOC levels at the experimental site were lower than the mean concentrations at traffic junctions. Particularly among nine VOCs, ethene (13.67 ± 4.12 ppbv), propylene and (14.4 ± 3.56 ppbv) were significantly higher by 4.3 and 4.2 times, respectively, at traffic junctions compared to TIFR site. These levels indicate the vehicular emissions, which are due to the internal combustion of engines. As the experimental site was located away from these traffic junctions, dissipation of pollutants to the atmosphere during transport of air mass led to relatively low concentrations (Swamy et al. 2012). So and Wang 2004 reported that the presence of the surrounding stature buildings can cause to prevent the vehicular emissions from dispersing effectively in urban areas.

Fig. 6
figure 6

The annual mean of hydrocarbon levels at different sites in the year 2013

Full size image

Ozone formation potential of VOCs

The important process of forming O3 in the lower troposphere is the photolysis of NO2 which is led by photochemical oxidation of hydrocarbons. Oxidation of hydrocarbons initiates by reacting with hydroxyl radicals in the atmosphere. Even though many reactions are involved in the formation of O3, the major processes can be summarized as follows:

$$ \mathrm{VOCs}+{\mathrm{OH}}^{\cdotp}\to {{\mathrm{RO}}_2}^{\cdotp }+\mathrm{products} $$
(8)
$$ {{\mathrm{RO}}_2}^{\cdotp }+\mathrm{NO}\to {\mathrm{NO}}_2+{\mathrm{RO}}^{\cdotp } $$
(9)
$$ {\mathrm{NO}}_2+{\mathrm{O}}_2\to {\mathrm{O}}_3+\mathrm{NO} $$
(10)

The rate of O3 formation depends on the amount of VOCs and NO x present in the atmosphere, rate constants for the VOCs initial reactions and level of OH˙ radicals (Carter 1994). To approximate the contribution of each hydrocarbon on the formation of O3, MIR values and propy-equiv. concentrations were introduced which do not consider the meteorological conditions (Carter 1994; Chameides et al. 1992; So and Wang 2004). Propy-equiv. is the concentration of hydrocarbon species on an OH-reactivity-based scale normalized to the reactivity of propylene. It represents the concentration (ppbC) of propylene required to yield a carbon oxidation rate equal to that of hydrocarbon species (Chameides et al. 1992). Ozone formation potentials were calculated by using maximum incremental reactivity. In these MIR scenarios, the NO x adjustment was done, where emission of reactive organic gases has the greatest effect on O3 formation, and NO x has the strongest O3-inhibiting effect (Carter 1994). The O3 formation potential is the product of the concentration of VOCs (μg/m3) and the MIR coefficient. In this study, propy-equiv. and OFP were estimated to predict the contribution of each hydrocarbon towards the formation of O3, where in below meteorological conditions and transport were not considered.

$$ \mathrm{O}\mathrm{F}\mathrm{P}(i)=\mathrm{concentration}(i)\times \mathrm{M}\mathrm{I}\mathrm{R}\;\mathrm{coefficient}(i) $$
(11)

Where concentration (i) = concentration of hydrocarbon i in μg/m3 and MIR = maximum incremental reactivity in grams of O3 produced per gram of VOC.

$$ \mathrm{Propy}-\mathrm{Equiv}(j)=\mathrm{concentration}(j)\times {k}_{\mathrm{OH}}(j)/{k}_{\mathrm{OH}}\left(\mathrm{propylene}\right) $$
(12)

Where concentration (j) = concentration of hydrocarbon j in ppbC, k OH (j) = rate constants for the reaction of hydrocarbon j with OH, and k OH (propylene) = rate constants for the reaction of propylene with OH. The rate constants of VOCs reaction with OH˙ at 298 K (cm3/molecule/s) were taken from published literature (Atkinson 1986, 1997; Atkinson and Arey 2003).

The annual mean propy-equiv. and OFP for TIFR site are listed in Table 1. Annual mean of propy-equiv. and OFP showed similar trend with the highest prop-equiv and OFP for propene, whereas ethane had the lowest values. Annual mean of propy-equiv. ranked in increasing order are as follows: ethane (0.07 ppbC), acetylene (0.38 ppbC), propane (0.56 ppbC), n-butane (0.87 ppbC), i-butane (0.97 ppbC), ethylene (2.02 ppbC), n-pentane (3.17 ppbC), i-pentane (3.96 ppbC), and propylene (6.07 ppbC). Annual mean of propylene with the highest propy-equiv. (6.07 ppbC) contributes to 34 % of the total VOCs propy-equiv., whereas acetylene and ethane with the lowest propy-equiv. contribute 2.1 and 0.38 % of the total VOCs, respectively. The highest propy-equiv. was observed for propene at our site due to its highest reactivity among these gases. The annual mean OFP of hydrocarbons measured are ranked in the following increasing order: ethane (1.17 μg/m3), acetylene (3.20 μg/m3), propane (4.06 μg/m3), n-butane(5.94 μg/m3), i-butane (8.58 μg/m3), n-pentane (14.14 μg/m3), i-pentane (23.72 μg/m3), ethylene (28.89 μg/m3), and propylene (35.67 μg/m3). Propene has the leading contribution of about 28.4 % of OFP, whereas ethane has the lowest of about 1 %.

Among the total hydrocarbons, the contribution of alkenes (only ethene and propylene) was 45 and 51 % for propy-equiv. and OFP, respectively. However, the contribution of six alkanes is about 53 and 46 %, and alkyne (acetylene) is 2 and 3 % for propy-equiv. and OFP, respectively. The high concentrations of propy-equiv. and OFP for alkenes are due to their high reactivity with hydroxyl radicals towards formation of O3. Among the six alkanes measured at our site, pentanes have the high propy-equiv. and OFP values due to higher number of reactive carbons in long chain molecules. The correlation for daily mean concentrations between O3 vs. propy-equiv. and O3 vs. OFP are presented in Fig. 7 for TIFR site. Scatter plots showed strong correlation of O3 with propy-equiv. and OFP with coefficients of determination about 0.65 and 0.77, respectively. The strong positive correlations indicate that O3 levels are related to VOCs chemical reactivity in the troposphere.

Fig. 7
figure 7

The correlation between a O3 vs. propy-equiv. and b O3 vs. OFP for daily mean concentrations in the year 2013

Full size image

Impact of VOCs on surface O3 levels as a function of NO x

As discussed above, O3 can be formed by the photochemical oxidations of VOCs in presence of sufficient NO x in the atmosphere. However, the destruction of O3 has been observed with increasing NO x concentration, particularly in the urban lower tropospheric region (Yerramsetti et al. 2013a). Relationship between NO x and VOCs on the formation of O3 revealed that reduced VOCs/NO x ratios can enhance titration reaction of O3 (Swamy et al. 2012). In this paper, an attempt was made to compute the higher limits of NO x for O3 reduction in the atmosphere in presence of various hydrocarbons during different seasons. The variations of surface O3 levels during daytime as a function of NO x and mean of VOCs during different seasons at TIFR site is shown in Fig. 8. It is understood that during all seasons, the increased NO x led to enhancement of O3 for a given VOCs concentrations at lower level of NO x . However, O3 reduction was observed after surpassing the crossover point at high levels of NO x . Similarly, at a given NO x concentration, with increasing VOCs concentration, O3 formation rate was observed to be increased at lower levels of NO x . However, there was no much formation rate of O3 at higher levels of NO x after surpassing the crossover point level. Increased titration of O3 with NO x was the possible reason for the reduction of O3 at high levels of NO x . Among the three seasons, high crossover value of NO x was observed during summer (9.5 ppbv) compared to monsoon (6 ppbv) and winter (9 ppbv). The high crossover point during summer was due to the most favorable conditions for photochemical formation of O3. Hence, high reaction rate of photochemical oxidation at elevated temperatures can control the titration of O3 even at high values of NO x .

Fig. 8
figure 8

Variations in surface O3 levels during daytime as a function of NO x and mean of VOCs during different seasons in the year 2013

Full size image

The intercomparison between hydrocarbons was also made to know the individual contribution to the crossover points. The variation in surface O3 levels during daytime as a function of NO x and different VOCs at TIFR site for all three seasons in the year is shown in Fig. 9. As discussed above, with increasing NO x concentration, there was enhancement of O3 at constant VOC concentrations at lower level of NO x with nine VOCs. However, decrease in O3 was observed after surpassing the crossover point at high levels of NO x . Similarly, at lower levels of NO x with increasing VOCs concentrations, O3 formation rate was observed to be higher, whereas no much formation rates of O3 observed after surpassing the crossover point level. The approximate crossover points of nine hydrocarbons are listed in the increasing order such as 6.5 (ethane), 6.5 (acetylene), 7.5 (propane), 8 (i-butane), 8 (n-butane), 8.5 (i-pentane), 8.5 (n-pentane), 8.5 (ethylene), 10 (propylene). Propylene showed the highest crossover point among hydrocarbons, whereas ethane and acetylene are the lowest (6.5). Strong correlation between crossover points vs. propy-equiv. and OFP are observed to be 0.78 and 0.74, respectively (Fig. 10). The crossover points are highly dependent on propy-equiv. and OFP values which are calculated from reactive rates of VOCs and concentration levels of such hydrocarbons. Thus, propylene has the highest crossover point among the hydrocarbons due to its high rate of photochemical oxidation, playing a role to form O3 even at high concentrations of NO x compared to other hydrocarbons.

Fig. 9
figure 9

Variations in surface O3 levels during daytime as a function of NO x and different VOCs in the year 2013

Full size image
Fig. 10
figure 10

Correlations between crossover points vs. propy-equiv. and O3 formation potentials at TIFR site

Full size image

Conclusions

Diurnal and seasonal patterns of VOCs along with NO x and the role of their chemistry on O3 variations have been discussed using ground data measurement at Hyderabad for the year 2013. Increased mixing boundary layer height and photochemical oxidation reaction during noon time are the dominant factors for low levels of VOCs as compared to morning and late evening. High diurnal amplitudes of ethylene and propylene were observed due to high reactivity of these alkenes with hydroxyl radicals during daytime compared to other hydrocarbons. Propy-equiv. and OFP indicates impact of hydrocarbon’s reactivity on the formation of O3 irrespective of their concentrations. Though the concentration of acetylene was high, the low propy-equiv. and OFP values were indicative of less reactivity towards formation of O3. The high crossover points of alkenes compared to alkanes also substantiated the formation of O3 even at high concentrations of NO x .