Volatile Organic Compounds in Roadside Environment of Hong Kong

Vehicular exhaust emissions are one of major sources of anthropogenic volatile organic compounds (VOCs) in urban areas of Pearl River Delta Region (PRDR). Six types of vehicle emission (VE)-dominated samples were collected at representative locations in Hong Kong in the winter of 2003. A total of 111 VOC species were quantified in the samples collected. n-Butane (31%) was the most abundant species in liquefied petroleum gas (LPG)-fueled VE-dominated samples, followed by propane (26%) and i-butane (25%). Toluene was the most abundant species in gasoline-fueled VE-dominated samples (16%), comprising about half of the quantified aromatic content. While ethene and ethyne have the greatest abundance in all diesel-fueled VE-dominated VOCs profiles (except at Tuen Mun Bus Depot). VOCs were also quantified at three roadside locations in Hong Kong. And ethene was the most abundant VOCs at roadside locations which accounted for 9.5 to 29% of the total quantified VOCs, except at Hong Kong Polytechnic University roadside monitoring station (PUX). Moreover, several VOCs were clearly in abundances in the roadside samples, namely toluene, ethyne, propane, ibutane, n-butane and i-pentane. Generally, strong and fair correlations were determined from the marker species of fuel vapor (i.e., LPG, gasoline, and diesel), which show significant fuel evaporation from vehicles in roadside environment of Hong Kong. Maximum incremental reactivity (MIR) was also calculated to evaluate the contributions of individual VOCs to ozone (O3) formation potential. The largest contributors to O3 production at Mong Kok roadside station (MKX) and Lok Ma Chau roadside station (LMX) were toluene (17 and 15% of the measured VOC reactivity, respectively), ethene (14 and 17% of the measured VOC reactivity, respectively), and propene (7 and 8% of the measured VOC reactivity, respectively), indicating the important roles of alkenes and aromatics in the ambient O3 formation.


INTRODUCTION
Volatile organic compounds (VOCs) are an important group of organic compounds in atmospheric chemistry.They can be either emitted into or formed in the atmosphere.Primary emission VOC sources include both natural (e.g., biomass burning) and anthropogenic (e.g., vehicular exhaust, cooking emission, incense burning, and cigarette smoke).Many VOCs have adverse health effects and play important roles in stratospheric ozone (O 3 ) depletion, the formation of toxic secondary organic aerosols, the formation of tropospheric ozone, and the escalation of global warming (Luis et al., 2003;Han and Naeher, 2006;Ran et al., 2009;Choi et al., 2011).Therefore, understanding source profiles of VOCs from vehicles (which is one of the major VOC sources in the urban areas) is critical to identify effective strategies for reducing VOC levels in the urban atmosphere.As one of the fastest growth regions in China, Pearl River Delta Region (PRDR) is experiencing severe air pollution due to rapid urbanization and increased use of motorized vehicles.Vehicular emissions (VE) are one of major sources of anthropogenic VOCs in the urban areas throughout the PRDR.Fuel reformation, use of lubricating oils, exhaust aftertreatment, and engine operation can highly vary the compositions and volatilities of the pollutants emitted into the atmosphere (Ålander et al., 2004).Guo et al. (2006) concluded from receptor model analysis that VE is a dominant VOCs source in Hong Kong, accounted for 39% of total emission, while it also accounted for 32% of total VOC emission in the PRDR.
Hong Kong is a densely populated city.According to the Hong Kong Transportation Department, there were 532,872 licensed vehicles in December 2004 (Hong Kong Transport Department, 2004).Several ambient studies have recently been completed indicating VE are the most important source for VOCs in Hong Kong (Ho and Lee, 2002;Lee et al., 2002;Guo et al., 2004).Liquefied petroleum gas (LPG), gasoline, and diesel are the main fuels used by vehicles in Hong Kong.In 2004, gasoline fueled vehicles accounted for 70.4% of the total licensed vehicles, while diesel and LPG fueled vehicles accounted for 24.5% and 3.5%, respectively (Hong Kong Transport Department, 2004).Vehicular performance affects fuel consumption and emissions in part because it would affect the combustion efficiency and evaporative emissions from the fuel system.In 2000, the Hong Kong Government implemented a subsidy program to switch diesel taxis to LPG taxis in order to reduce the impact of smoke, NO x and VOC emissions on the street level air pollution in Hong Kong (Hong Kong Environmental Protection Department, 2001).Nearly all (99.8%) of the 18,000 taxis in Hong Kong were LPG powered in 2003.In PRDR, most previous studies (e.g., Lee et al., 2002;Ho et al., 2004;Chan et al., 2006;Guo et al., 2007) reported the urban/rural levels of VOCs.To our best knowledge, there are only limited measurement-based VOC emission profiles available in Hong Kong (Tsai et al., 2006;Ho et al., 2009;Guo et al., 2011;Ho et al., 2012).In previous studies, we have determined the emission factors of VOCs at a roadway tunnel in Hong Kong (Ho et al. 2007(Ho et al. , 2009)).Influences from VE to the local ambient VOCs in roadside environments had not been thoroughly investigated.In this study, each type of VE-dominated samples was collected in the atmosphere in a location (e.g., taxi stations and bus depots) where is mainly contributed by a particular emission from vehicles fueled with liquefied petroleum gas (LPG), gasoline, or diesel.The VOC concentrations and profiles of the source dominated samples were compared.Samples were also collected at three representative roadside sites.The VOC markers for LPG, gasoline, or diesel can be used as indicators for source identification in the roadside environments.The in-depth understanding of source contributions provides important information for management of Hong Kong air quality.

Sampling Sites
Sample collection can be divided into two different categories including (I) VE-dominated and (II) roadside samples (Fig. 1 and Table1).
(I) VE-dominated samples: In order to obtain the comprehensive chemical source profiles for different type of VE-dominated emissions in Hong Kong, ground-based roadside sampling at six traffic intersections and highway on/off ramps was conducted.The most important use of vehicle emission dominated profiles is for the source marker identification for the different types of vehicular exhaust.The six sampling sites are described as below: 1) The Hong Kong Polytechnic University (HKPU) car park station (PCX) is situated at ground floor of the HKPU, Hung Hom.The sampling equipment was placed at the entrance of the car park.It was chosen to represent the gasoline-fueled light-duty passenger car emissions.2) Wan Chai liquefied petroleum gas (LPG) refilling station (TAS) is located at Wan Shing Street, Wan Chai.The sampling equipment was placed at the roadside near the LPG refilling station.It was chosen to represent the LPG taxi emission.
3) Shau Kei Wan minibus station (MBS) is situated at Mong Lung Street, Shau Kei Wan.The sampling equipment was placed next to the minibus station.It was chosen to represent the minibus diesel emission.4) Cheung Sha Wan Whole Food Market station (GVX) is located at the Yen Chow Street West, Cheung Sha Wan.The sampling equipment was placed at the entrance of the whole food market.It was chosen to represent the light-and heavyduty vehicle diesel emission.5) The Peak station (SBS) is situated at the Peak car park, near the Peak Road.The sampling equipment was placed at the roadside of the car park.It was chosen to represent the single-decked bus diesel emission.6) Tuen Mun Bus Depot station (DBS) is located at the Ho Tin Street, Tuen Mun.The sampling equipment was placed at the roadside of the bus depot.It was chosen to represent the double-decked bus diesel emission.
(II) Roadside samples (collected in the summer and winter of 2003): 1) The Mong Kok Environmental Protection Department (HKEPD) roadside air quality monitoring station (MKX) is located at the junction of the heavily traffic Lai Chi Kok and Nathan Roads.2) HKPU roadside monitoring station (PUX) is located 1 m adjacent to Hong Chong Road, which leads to the Cross Harbor Tunnel.It can represent street-level emissions from nearby vehicle exhaust and road dust.It is aimed to get slow traffic emission exhausts when vehicles are idling and under slow traffic/for crossboundary vehicles.3) Lok Ma Chau roadside monitoring station (LMX) is located at the entrance of the boundarycrossing point that vehicles come from Mainland China to Hong Kong.

Engine and Fuel Types
There were a large number of pre-Euro diesel-fueled buses and heavy duty vehicles on the road in Hong Kong.In 2003, the pre-Euro fleet contributed > 40% of the total number of heavy duty vehicles and > 10% of the buses.And Euro II fleet also had a contribution of > 24% for heavy duty vehicles and > 30% for the buses.In this study, classification of the diesel-fueled engine type by the European emission standard was not done owing to too many vehicles passing the monitoring stations in a short period.For the same reason, classification of gasoline vehicles by either models or years was not applicable as well.Therefore the inter-variations of engines for each fueled type were not taken into account.Euro III unleaded

Sample Collection
A total of 51 VE-dominated and roadside samples were collected in Hong Kong between 8:00 and 19:00 in summer (23 May-9 July) and winter (9 December-28 December) of 2003.At least three samples were collected for each type of VE-dominated samples while twelve samples were collected in each roadside station.Ambient volatile organic canister samplers (AVOCS) (Andersen Instruments Inc. Series 97-300, Smyrna, GA, USA) were used to collect whole air roadside samples into pre-cleaned and preevacuated 2-L stainless steel canisters at a flow rate of 20 to 25 mL/min for 3 h in winter and 2 h in summer, respectively.The canisters were pressurized when sampling.
The sampler was fixed on the ground level with an inlet at a height of ~1.5 m.The flow rates were checked in the field before and after each run using a calibrated flow meter.After sampling, the filled canisters were shipped to the laboratory of the University of California, Irvine for chemical analysis within two weeks of being collected.

Sample Analyses
All canisters were shipped to the laboratory at the University of California, Irvine (UCI) and analyzed for carbon monoxide (CO), carbon dioxide (CO 2 ), methane (CH 4 ), and non-methane hydrocarbon (NMHCs).CO and CO 2 analyses were carried out using a hydrogen gas methanizer upstream of a HP 5890 gas chromatography (GC) equipped with a flame ionization detector (FID) and an 3m molecular sieve column.CH 4 was also analyzed using an HP 5890 GC equipped with an FID.The samples were injected into an 1/8" stainless steel 0.9 m column packed with 80/100 mesh Spherocarb.The analytical system used to analyze NMHCs (i.e., saturated, unsaturated, aromatic, and halogenated hydrocarbons) involved the cryogenic pre-concentration of 1520 ± 1 cm 3 (STP) of air sample in a stainless steel tube filled with glass beads (1/8" diameter) and immersed in liquid nitrogen (-196°C).A mass flow controller with a maximum allowed flow of 500 mL/min controlled the trapping process.The trace gases were revolatilized using a hot water bath and then reproducibly split into five streams directed to different column-detector combinations.

Photochemical Reactivity
It is well known that VOCs are significant precursors of O 3 formation (Chameides et al., 1992).Individual compound has different characteristic photochemical reactivity.In order to calculate the O 3 -forming potential of the VE, the speciated emission factors for each vehicle type were multiplied by the maximum incremental reactivity (MIR) scale developed by Carter (1994).The MIR are in units of grams of O 3 per gram of organic compound and therefore are simply multiplied by the emission factors (grams of organic compound per vehicle-km driven), to yield reactivityadjusted emission rates in units of O 3 per vehicle-km.

Vehicular Emission-Dominated Samples
Although some of the sources of VOCs in Hong Kong originate from fuel combustion, no comprehensive chemical source profiles for individual type of VE have been performed.There is limited information on source or source-dominated emissions in Hong Kong.Mobile source emissions are among the most difficult to measure with respect to chemical composition.This difficulty arises from: 1) different mobile source types; 2) inadequate characterization of the high emitters within motor vehicle fleet; 3) a large number of individual emitters within each vehicle category; 4) fuel use characteristics and emission control technology that change from year to year; 5) undefined operating conditions; 6) several emission points on each vehicle i.e. tailpipe and fuel evaporation; and 7) a mixture of primary particles, semi-volatile compounds, and secondary particle precursors (Watson et al., 1990).Source-dominated profiles for motor vehicle emissions were constructed from ground-based roadside sampling at various traffic intersections and highway on/off ramps where the sampled air was dominated by emissions from different type vehicle exhaust.The most important use of source-dominated profiles in this section is for the source marker identification for the six different types of vehicular exhaust which is very useful in source apportionment using receptor models.
In this study, a total of 111 species were quantified in the samples collected.These include CH 4 , CO, CO 2 , carbonyl sulfide (OCS), carbon disulfide (CS 2 ), methyl tertiary butyl ether (MTBE), 40 C 2 -C 10 saturated hydrocarbons, 32 C 2 -C 10 unsaturated hydrocarbons, 21 C 6 -C 10 aromatic hydrocarbons and 12 halogenated hydrocarbons.Table 2 shows the VOC concentrations collected from different VE-dominated locations in Hong Kong.It is critical to point out that the ambient levels and contributions of VOCs were possibly varied by other local anthropogenic emission sources, regional transportation of pollutants, and photochemical reactions of organic compounds in the atmospheres.For these reasons, the samplings were carried out at rush hours in order to obtain the most representative samples.
The highest total quantified VOC concentration (except CH 4 , CO and CO 2 ), 682 ppbv, was found in the gasolinefueled light-duty passenger car (PCX) sample.The LPGfueled taxis dominated samples were collected in TAS which had a mean concentration of 526 ppbv.The mean total concentrations were 140 and 76 ppbv, respectively, at the diesel-fueled minibus (MBS) and double-desk buses (DBS) stations.The total quantified VOC concentrations was 58 ppbv on average at the diesel-fueled single-desk buses station (SBS) while a lower value of 37 ppbv was found in the diesel-fueled goods samples (GVX).
Fig. 2 shows the variations of different groups of VOCs at the source-dominated roadside environment.Saturated hydrocarbons are the most abundant (> 85%) group in LPGfueled dominated samples because the major compositions of LPG in Hong Kong are butane and propane.Therefore, the emissions from LPG-fueled dominated sample are mainly   low molecular weight VOC species.In gasoline-fueled dominated samples, unsaturated hydrocarbons are the most abundant (36%), followed by unsaturated hydrocarbons (32%) and aromatic hydrocarbons (30%).The relatively high aromatic fraction in gasoline enhances its combustion performance but generally at the expense of an increase in the NO x emissions (Wang et al., 2001).And in diesel-fueled dominated samples (except minibus), saturated hydrocarbons are the most abundant, followed by unsaturated hydrocarbons and aromatic.Moreover, the abundance of halogenated hydrocarbons in diesel-fueled dominated samples are relatively higher than LPG-fueled and gasoline-fueled dominated samples.VOC molar composition profiles in the different types of VE-dominated samples are shown in Fig. 3.It shows that n-butane (31%) constituted the largest fraction of measured VOCs in TAS sample, followed by propane (26%) and i-butane (25%).These three species are the major constituents of LPG in Hong Kong (Tsai et al., 2006).
Toluene was the most abundant aromatic in PCX samples (16%), comprising about half of the quantified aromatic content.Toluene is a commonly used gasoline additive and the amount of toluene added varies with different oil companies.The second most abundance VOC specie is ethene (8.4%), followed by pentane (7.1%) and ethyne (6.6%) in PCX sample.Toluene and i-pentane were the two main constituents of gasoline and they were used as tracers for estimating gasoline evaporative losses (Tsai et al., 2006).However, in PCX samples, the abundance of MTBE is significantly higher (2.1%).MTBE is a gasoline additive, used as an oxygenate and to raise the octane number.According to previous studies, evaporative loss of unburned fuel is an important emission source from automobiles (Na et al., 2004).Generally speaking, ethene, and ethyne have the greatest abundance in all the other diesel-fueled dominated VOCs profiles (except DBS samples).They are the major sources for incomplete combustion.And in DBS sample, n-butane and toluene are the most abundance VOC species.They are the major sources for fuel evaporation.As diesel is a mixture of hydrocarbons with higher density, viscosity and sulfur content than gasoline or LPG, the abundances of OCS and CS2 (sulfur emission) in SBS, DBS and GVX samples (mainly diesel-fueled emission samples) are higher than in LPG-fueled and gasoline-fueled dominated samples.Anthropogenic sources of OCS arise from the combustion of biomass and fossil fuel.Emission of OCS from diesel-fueled vehicles is one of such example.

VOC Concentrations in Roadsides
Table 3 shows the VOC concentrations collected in the atmosphere at different roadside locations in Hong Kong.The vehicle numbers and their fuel types were counted during sampling events in order to evaluate their effects on the ambient concentrations.The statistical counting precision were < 1%.On average, the mean concentrations of total quantified VOCs (sum of all of species except for CH 4 , CO 2 , and CO) at Mong Kok (MKX), The Hong Kong Polytechnic University (PUX), Lok Ma Chau (LMX), were 74.9 ± 12.1 ppbv, 238 ± 117 ppbv, 58.5 ± 63.8 ppbv respectively.The concentration of total quantified VOCs at PUX was the highest, especially in summer, as the site is at the tunnel exit in downtown area.Traffic flow record shows an average of > 5,000 of different classes of LPG-, gasoline-, and diesel-fueled vehicles accessing the site during the sampling events (LPG: 21%; gasoline: 41%; and diesel: 38%).Moreover, there is a traffic light near the sampling location at PUX, emissions resulting from vehicle idling increased.In addition, the high concentration of C 6 to C 9 saturated hydrocarbons observed at PUX perhaps may be due to the evaporation of fuels from fuel tanks of gasoline-fueled vehicles or other special VOCs sources near the sampling site.Except the unknown saturated hydrocarbons emission, toluene has the greatest abundance in all the VOC profiles, this suggests that roadside areas are significantly influenced by the gasoline-fueled VE.The lowest (0.98 ppbv) and the highest (36.1 ppbv) mean concentrations of toluene were also observed at LMX (with high standard deviation), it means the emission of toluene in LMX was not stable during the sampling period.The site is the main transportation gate for goods between Hong Kong and China.More than 80% of the vehicles passing through the site were diesel-fueled light and heavy goods vehicles during the sampling events.Although toluene is the most abundance specie of gasoline evaporation but significant amount of toluene can be observed in diesel fuel also (Tsai et al., 2006).This indicated the importance   of running evaporative loss from the fuels in the roadside environment.Among the VOCs analyzed, the concentration of ethene (10 ppbv) was the most abundant specie in LMX, followed by toluene (5.9 ppbv) and ethyne (4.5 ppbv).
High concentrations of LPG tracers (n-butane, i-butane and propane) were also observed in roadside environment.This suggests that the downtown area is significantly influenced by the LPG-fueled VE.Statistic shows that the amounts and contributions of LPG-fueled taxi accessing PUX were higher than MKX, therefore, higher levels of LPG tracers were found in PUX than in MKX.Moreover, high concentrations of ethene and ethyne were observed.They are typical tracers for combustion, and thus VE was the likely source of these two compounds (Barrefors and Petersson, 1996;Stoeckenius et al., 2006).Fig. 4 shows the abundance of four groups of VOCs at each roadside sampling location.Saturated hydrocarbons are the most abundant group, followed by unsaturated hydrocarbons, and aromatic hydrocarbons in PUX and MKX.Moreover, the VOC profiles at roadside locations are shown in Fig. 5. Ethene was the most abundant VOCs which accounted for 9.5 to 29% of the total quantified VOCs, except in PUX (due to the unknown saturated hydrocarbons emission).Moreover, it is clear that there are several VOCs that were abundant in the roadside samples, namely toluene, ethyne, propane, i-butane, n-butane and i-pentane.Our findings were consistent with the compositions reported by Tsai et al. (2006) and Ho et al. (2009).The result also promises an important input of fuel from gasoline-fueled and LPGfueled vehicles to the roadside areas.As diesel samples consisted mainly of heavy C 8 -C 10 alkanes, these compounds have low vapor pressures and thus do not readily evaporate into the atmosphere, suggesting that evaporative loss from diesel to the roadside atmosphere was insignificant (Tsai et al., 2006).

VOC Ratios
A high ratio of a more reactive VOC to a less reactive VOC (photochemical lifetimes or reactivities) against hydroxyl radical (OH) indicates relatively little photochemical processing of the air mass and major impact from primary emissions.On the other hand, a lower ratio is reflective of more aged VOC mixes and thus presumably that the VOCs were emitted from more distant sources (Guo et al., 2007).Comparisons of the ratios can be used to estimate the relative ages of air parcels.In this study, the ratio of m,pxylene/ethylbenzene were used for comparison (Nelson and Quigley, 1983;Smyth et al., 1999;So and Wang, 2004).The average value of m,p-xylene/ethylbenzene in tunnel (2.61 ± 0.30) (Ho et al., 2009) are close to/or slightly higher than the ratios measured at the roadside sampling location at MKX, PUX, LMX (2.16 ± 0.32, 2.74 ± 0.22, 2.02 ± 0.55) and higher than other urban/rural sites in Hong Kong (range from 1.3-1.8)(Guo et al., 2007).Ethene and ethyne are typical tracers for combustion, and thus vehicle exhaust was the likely source of these two gases (Barrefors and Petersson, 1996;Stoeckenius et al., 2006).Tsai et al. (2006) concluded that the ethyne/ethene ratio for Hong Kong was 0.53 ± 0.03.And the studies conducted in highways tunnels throughout the U.S. indicate that the ratio range from 0.59-1.0(Harley et al., 2002).The average ethyne/ethene ratio in this study ranged from 0.32 to 1.04 which is close to the previous studies.These indicated that the three sites have more influence from VE.
The major sources of toluene and benzene in urban areas are VE and solvent use.Toluene has a shorter life time (~3 days) than benzene (~12 days).The average toluene/benzene (T/B) ratios were 22.1 ± 12.7 in the gasoline used in Hong Kong (Tsai et al., 2006).The T/B ratios in gasoline were higher than the ratios found at a roadside microenvironment in Hong Kong (ranged from 1.4 to 9.6; an average of 5.2 ± 2.2).Different source-dominated samples have different T/B ratios (range from 2.8 to 6.0), and the variations should depend on the fuel use characteristics, emission control technology and fuel evaporation.And the differences were mainly due to the variations of toluene emission.More studies should be done in order to have a clear picture for these source characteristics.

Correlation of VOCs
The correlations among CO and VOCs for all roadside samples were evaluated by correlation analysis.CO is generally emitted from incomplete combustion of fossil fuel.Determining the relationship between CO and VOCs can provide useful information on their sources and emission signatures (Wang et al., 2002(Wang et al., , 2003;;Guo et al., 2007).Among the VOCs measured, ethane was best correlated with CO (R = 0.94), follow by propane and ibutene (R = 0.83) as well as ethyne (R = 0.81) confirming a common source origin in roadside environment of Hong Kong.Generally, good and fair correlations were observed between most of the combustion emitted VOCs and CO, except halogenated hydrocarbons, isoprene and pinenes which are emitted from different sources.Isoprene is mainly emitted from biogenic sources (e.g., vegetables) although a few studies report the contribution of isoprene from VE (e.g., Borbon et al., 2001;Barletta et al., 2002).Here, the poor correlation between isoprene and CO suggests that VE of isoprene is not significant.Moreover, biogenic isoprene emissions from the vegetation near the sampling locations may have significant impact to the result.Moreover, strong and fair correlations were determined from marker species of fuel vapor (LPG, gasoline, and diesel).As observed in previous section, propane, n-butane and i-butane are the major constituents of LPG-fueled vehicles, strong correlations were found (R = 0.81-0.94)among the species which indicated that unburned LPG was emitted to the roadside atmosphere.n-pentane, i-pentane, toluene, m-xylene, pxylene, o-xylene are the most abundant VOCs from the exhaust of gasoline-fueled vehicle and the evaporative loss from gasoline vapor (Ho et al., 2009).Fair and strong correlations (R = 0.64-0.92) of these species indicated the importance of emissions from gasoline-fueled vehicles.Moreover, good correlations (R = 0.69-0.80)were observed among diesel-fueled species (n-octane, n-nonane and ndecane), which is match with the results of Tsai et al. (2006).

Reactivity with Respect to Ozone Formation
It is well known that VOCs are significant precursors of O 3 formation.Individual VOCs have different photochemical reactivities.To assess the relative importance of VOCs from roadside environment in Hong Kong, we have applied the MIR to evaluate the contributions of individual VOCs (except OCS, H-1211, CHBr 3 , and isobutylbenzene, as no MIR values are available for those species) to O 3 production (Carter, 2010).The product of the concentration and the MIR coefficient, where MIR is the amount (g) of O 3 formed per gram of VOC, indicates how much the individual VOCs may contribute to O 3 formation in the air mass (Grosjean et al., 1998).Although the calculations of O 3 formation in this study are not very accurate, it provides rough ideas about individual VOCs on their photochemical reactivity.
The top ten reactivities with respect to O 3 formation of three roadside environments are presented in Table 4.The contributions of these ten VOCs were 51 to 64% of those of total measured VOC reactivity.The largest contributors to O 3 production in MKX and LMX were toluene (17 and 15% of the measured VOC reactivity, respectively), ethene (14 and 17% of the measured VOC reactivity, respectively), and propene (7 and 8% of the measured VOC reactivity, respectively), indicating the importance of alkenes and aromatics in the production of O 3 .However, in PUX, the largest contributors to O 3 production were toluene (9% of the measured VOC), 3-methylhexane (6% of the measured VOC) and n-hexane (6% of the measured VOC).Toluene, 3-methylhexane and n-hexane were mainly from sources such as gasoline evaporation and use of solvents.Therefore, other than VE, fuel evaporation (which depends on temperature) also makes contributions to the O 3 production in roadside environment of Hong Kong.

CONCLUSIONS
In order to study the chemical concentrations and profiles of the source dominated samples, six types of VE-dominated samples were collected at representative locations in Hong Kong.A total of 111 VOCs were analyzed and reported.The dominant species for LPG-fueled dominated samples were n-butane, followed by propane (26%) and i-butane (25%).Toluene (16%), i-ethene (8.4%), pentane (7.1%) and ethyne (6.6%) were the dominant species for gasoline-fueled dominated samples while ethene and ethyne were the most dominant species for diesel-fueled dominated samples (except in DBS samples).Moreover, the abundances of OCS and CS2 (sulfur emission) in diesel-fueled dominated samples are higher than in LPG-fueled and gasoline-fueled dominated samples because of higher sulfur content in diesel than gasoline and LPG.Except the unidentified sources for saturated hydrocarbons at PUX, toluene has the highest abundance in all the roadside sites, suggesting that roadside areas are significantly influenced by the gasolinefueled VE in Hong Kong.This gas acts as a tracer of gasoline evaporation and its high abundances indicates the importance of running evaporative loss from gasoline-fueled vehicles in the roadside environment.The strong and fair correlations were determined from the marker species of fuel vapor (i.e., LPG, gasoline, and diesel), which show significant fuel evaporation from vehicles.The O 3 formation potential (OFP) of individual VOCs for VE were assessed.The largest contributors to O 3 production in roadside environments were toluene, ethane, propene, methylhexane and n-hexane, indicating vehicular fuel evaporation also has contributions to the O 3 production in roadside environment of Hong Kong.The in-depth understanding of source contributions provides important information for management of Hong Kong air quality.

Fig. 1 .
Fig. 1.The map showing the sampling locations in this study.Location acronyms are shown in Table1.

Fig. 2 .
Fig. 2. The abundance of five groups of VOCs for the source-dominated samples.

Fig. 3 .
Fig. 3. Average VOC molar composition profiles for the vehicular emission dominated samples.

Fig. 4 .
Fig. 4. The abundance of five groups of VOCs for the roadside samples.

Table 1 .
Brief summary of the sampling locations of vehicle emission dominated samples.

Table 2 .
VOC and other gaseous pollutant concentrations in different types of vehicular-emission-dominated samples.

Table 3 .
VOC and other gaseous pollutant concentrations in three roadside locations in Hong Kong.

Table 4 .
Top 10 VOCs for ozone-forming potential at the three roadside locations.
a Percent of total O 3 formed by carbonyls.