Traceability tagging of volatile organic compound sources and their contributions to ozone formation in Suzhou using vehicle-based portable single-photon ionization mass spectrometry

Background In recent decades, there has been an increasing global preoccupation with atmospheric volatile organic compounds (VOCs). Given the significant impact of VOCs as pollutants and essential precursors of ozone (O 3 ) in urban and industrial areas, it is imperative to identify and quantify the sources of their emissions to facilitate the development and implementation of effective environmental control strategies. Methods A mobile laboratory vehicle equipped with a single-photon ionization–time-of-flight mass spectrometer (SPI–TOFMS) and a navigation system was employed to establish the traceability of VOCs that contribute to the formation of ozone in Suzhou Industrial Park. The method exhibited a favorable detection limit of 0.090

has demonstrated a promising methodology that is well-suited for the real-time and online monitoring of VOCs in the atmosphere.In addition, a library for identifying VOC fingerprints from the same plant was established.This library serves as a comprehensive resource for establishing on-site VOC traceability, estimating source apportionment, and evaluating their impact on ozone formation.Keywords Volatile organic compounds, Single-photon ionization, Time of flight, Mass spectrometry, Positive matrix factorization, VOC fingerprint

Background
The swift industrialization and urbanization in China have had significant implications for air quality and ecological stability and have led to notable challenges in environmental pollution [1][2][3][4].The emission of volatile organic compounds (VOCs) into the atmosphere is undergoing rapid and exponential growth, emerging as a significant contributor to atmospheric pollution in China [5,6].In recent years, numerous research studies have been conducted throughout China to investigate the sources of VOCs emissions [7][8][9].A significant proportion of VOCs consist of anthropogenic chemicals utilized in various manufacturing industries, such as paints, pharmaceuticals, and refrigerants.These compounds also originate from vehicle emissions [2,3].A notable proportion of VOCs are known for their unpleasant odors and present significant health hazards, leading to both acute and chronic illnesses, including respiratory diseases, premature mortality, lung cancer, and childhood asthma [10][11][12][13][14][15][16].
VOCs play a crucial role in atmospheric photochemical processes as key precursors to O 3 formation and other secondary organic aerosols [17].Research investigating the correlation between O 3 and its precursors suggests that the formation of O 3 is limited in urban as well as industrial areas [18,19].Industrial-related emissions of VOCs, including benzene, toluene, and xylenes, prominently contribute to O 3 formation and the generation of secondary organic aerosols [20].Tracing, identifying, and quantifying the sources of VOCs and other pollutant emissions are essential prerequisites for the analysis and development of strategies for controlling air pollution.Consequently, the development of an effective technique for managing the photochemical process, tracing VOCs emission sources, and quantifying their relationship with O 3 formation becomes imperative.
The reactivity-based approach, as proposed by [21], combines the positive matrix factorization (PMF) model with an observation-based model (OBM).This methodology was exclusively applied within the confines of nonurban interior zones.Since its establishment, the PMF model has been widely utilized for the purpose of identifying and quantifying VOCs in various regions across the globe [21][22][23].
In the present study, the PMF model is employed to delineate the primary sources of VOCs and to assess their contributions to the photochemical formation of O 3 .The OBM, grounded in the carbon bond 4 mechanism, provides a valuable tool for scrutinizing the relation between O 3 and its precursors at specific geographical locations.These precursor compounds include VOCs, nitric oxide (NO x ), and carbon monoxide (CO) [23][24][25].Unlike models that rely on emissions, the OBM incorporates observed ambient concentrations of O 3 and its precursors to simulate the photochemical formation and destruction of O 3 .This approach helps to mitigate uncertainties associated with emission inventories [21,26,27].
However, some researchers tend to emphasize the contribution of specific VOCs to O 3 formation, often neglecting the influence of others.Such an approach may lead to inadequate analysis of the data.To overcome this, we adopt a comprehensive stance, considering each individual VOC, and study the source contribution to the O 3 formation.
Various analytical techniques have been documented in the existing literature to identify and trace the source of VOCs in the atmosphere.These analytical techniques encompass various methods, such as spectroscopy and inductively coupled plasma-mass spectrometry (ICP-MS) [28], laser-induced breakdown spectroscopy, portable modular biological mass spectrometry (MoBiMS) [29], proton transfer reaction-mass spectrometry (PTR-MS) [30,31], and membrane inlet-mass spectrometry (MI-MS) [32].While these traditional techniques are used for monitoring both atmospheric organic and inorganic volatile compounds, as well as volatile organohalogen compounds (VOHCs) in the gaseous phase [29,[33][34][35][36][37].However, their scope of application is constrained due to inherent limitations such as poor quantification capacity, limited detection range, intricate and costly procedures, and the need for highly skilled personnel [29,38].Consequently, there exists a compelling demand for the development of a rapid and straightforward analytical method capable of furnishing a standardized and accurate approach to tracing VOCs sources in a given area of interest across all concentration levels in the atmospheric environment.This technique aims to overcome the above limitations observed in VOCs analysis, improve the detection capability, and minimize the analysis time of VOCs.This provides an effective way of monitoring the air quality of the atmospheric environment.
In this study, we utilized a mobile laboratory navigation vehicle mounted with a portable self-developed device based on time-of-flight mass spectrometry (TOFMS) coupled with vacuum ultraviolet-single-photon ionization (VUV-SPI).This configuration facilitates real-time detection and monitoring of atmospheric pollution while concurrently tracing VOCs emission sources across the entire concentration range.In addition, we compile a comprehensive VOCs fingerprint library for distinct sources in Suzhou, Jiangsu Province, China.The ease of spectral interpretation is enhanced through the use of hyperspectral imaging, soft ionization sources, and the ability to ionize molecular ions, particularly for analytes in complex matrices [39].TOFMS has been extensively utilized in the field of environmental air monitoring and inspection of industrial pollutions, as evidenced by its widespread application [40][41][42].The method demonstrated excellent performance besides its good detection and quantification limits (LOD and LOQ) and a small range of relative standard deviation (RSD).

Vehicle platform
The Puwei AQM-100 atmospheric navigation vehicle, modified by Suzhou Weimu Intelligent System Co., Ltd., utilizes an IVECO car from Navico, Ltd., Nanjing, China, for online atmospheric air monitoring.This platform is capable of capturing real-time concentration distributions of diverse VOCs.The system is equipped with various instruments to facilitate continuous monitoring operations for 24 h.These include an SPI-MS instrument (TMS-400) and a mass analyzer, a meteorological monitoring sensor (PSG04175 Leeds, LS12 2QQ, England), and an uninterrupted 6 kW high-power supply generator (DXGN6000, EM POWER, Florida US) are installed.Figure 1 illustrates the setup.Furthermore, a global positioning system (GPS receiver) (NV08C-CSM, Montlingen, Switzerland) to provide geographical data.The meteorological sensor is responsible for monitoring various meteorological parameters, including wind speed and direction, temperature, humidity levels, and atmospheric pressure.Real-time air pollution data are visualized on an electronic map through the utilization of the Air Scan System Software.
The vehicle comprised three distinct components: a front driver seat area similar to standard vehicles, a middle operating room that houses the GPS receiver for the purpose of monitoring geographical data, and a rear instrument compartment.

Experimental site
In this study, the geographical locations of the sampling sites are depicted in Fig. 2. The sites were located 100 km east of Shanghai and 200 km from Nanjing, with an average elevation ranging from 3.5 to 5 m above sea level.The sampling sites experience a subtropical monsoon climate characterized by four distinct seasons, similar to the climates of Shanghai, Hangzhou, and Nanjing cities [9].To understand the spatial distribution and seasonal variations of VOCs concentration in Suzhou, Jiangsu Province, China (31° 18′ 14.69″ north latitude and 120° 35′ 43.37″ east longitude), the details of the sampling navigation routes are shown in Additional file 1: Fig. S1a, b.The monitoring experiment on atmospheric VOCs was conducted using an online mobile laboratory instrument during morning rush hours (8:00-9:00 a.m.) and in the evening (5:30-6:30 p.m.) from February 2022 to January 2023 along the industrial park.In this study, the overall average atmospheric VOCs concentration was determined by calculating the average concentration for each sampling period under specific weather conditions.Six samples were collected each for winter (December to January), and summer (July to August), both during morning rush hours and in the evening.The average concentration was computed for each category as follows: the average concentration of VOCs for each of the four subgroups, namely, morning rush hours and evening during winter and summer.Subsequently, after obtaining these subgroup averages, the overall average concentrations during the monitoring period were determined.

SPI-TOFMS instrument
In the present study, a self-made integrated single-photon ionization-time-of-flight mass spectrometry (SPI-TOFMS) was employed to trace the source of VOCs.The system consists of several components, including a membrane inlet, an ionization system, a data acquisition system, a vacuum system, an electrical system, and a reflection TOFMS.The schematic diagram of the instrument is illustrated in Fig. 3.The dimensions of the SPI-TOFMS instrument are 50 cm × 50 cm × 50 cm, and it has an approximate weight of 50 kg.The device is equipped with a battery and a mechanical pump, allowing for its utilization in vehicles and mobile operations.
The pumping system comprises a scroll dry pump (SH110, Varian, Palo Alto, CA, USA) with a flow rate of 1 Lmin −1 and a molecular turbo pump (V301, Varian, Palo Alto, CA, USA) with an operating speed of 300 Ls −1 .This ensures the requisite air pressure and optimal functionality.The dry pump operates at a pressure of approximately 200 Pa, which is supplied by the primary vacuum system.In addition, the molecular pump achieves a vacuum level that exceeds 2.1 × 10 -3 Pa, thereby meeting the analytical specifications.The instrument demonstrates a rapid response speed in fractions of seconds and is capable of detecting 10 spectra per second.In addition, the SPI-TOFMS instrument is equipped with a high-speed data acquisition card, specifically the 1 GS/s Keysight U5309 (Keysight Technologies, USA), which enhances the efficiency of data processing and enables an extended collection period.

Membrane inlet system
The portable TOFMS is equipped with an inlet system consisting of a membrane inlet (MI) and a capillary inlet (CI).In the present study, the VOCs were gathered and underwent filtration using a silicon sheet membrane made of 90% polydimethylsiloxane (PDMS) from (Technical Production, Inc., St. Louis, MO, USA).The membrane had an effective diameter of 20 mm and a thickness of 50 µm.VOCs were introduced into the ionization region using a vacuum pump with a pumping speed of 1 Lmin −1 .

Single-photon ionization (SPI)
The SPI system comprises essential components, including a vacuum ultraviolet (VUV) lamp, an ionization chamber, a derivation electrode, a focus electrode, an output electrode, and a lens, as shown in Fig. 3. VUV lamps (low-pressure Krypton discharge lamps) are sourced from MgF 2 (Heraeus, Germany), are used to The transmission path is sealed and kept short to maintain high gas sample flow at 1000 mLmin −1 , allowing for more ionization to occur.When the VUV lamp is ignited with a voltage of 1100 V, the intensity of ions produced is directly proportional to the current flowing through the lamp and it can be regulated to a value of 1 mA.Lenses play a crucial role in the optimization of ion beam performance by facilitating the concentration of ions, thereby improving ion transmission efficiency, enhancing detection signal sensitivity, and mitigating the risk of mechanical misalignment.

Time-of-flight mass spectrometer (TOFMS)
The configuration of the SPI-TOFMS system consists of several distinct components, including a doublepulse acceleration region, a drift tube, a reflector, and an ion detector, as illustrated in Fig. 3.The development of TOFMS was initially based on an apparatus designed by Dodonov [43].When ions with the same energy level but different mass-to-charge ratios traveling at a specific speed, they experience uniform acceleration in an electrostatic field with the measurement scan frequency of 10 kHz.This acceleration results in the ions having the same kinetic energy, and they subsequently reach the detector by passing through the drift tube.Lighter ions arrive earlier than the heavier ions, and their spectra are subsequently recorded.Determining the time of flight of each ion facilitates subsequent analysis.Peak integration and data acquisition were conducted using the Xcalibur ® software (version 2.2 SP1.48; TFS, San Jose, CA, USA), which is embedded in the instrument system.The duration of each scan is approximately 100 ms, as determined by the experimental conditions, and can be adjusted to meet specific requirements.

Positive matrix factorization (PMF) model
Positive matrix factorization (PMF) is a robust multivariate factor analysis technique used to decompose a matrix of sample-specific data into two matrices: factor contributions and factor distribution.An analyst conducts an analysis of the matrices to ascertain the sources that are represented, using observations from receptor sites [44].In this study, Eq. 1 depicts the concentration of the jth chemical species in the ith samples, which is influenced by p independent sources: where x ij is the jth species concentration measured data matrix in the ith sample variable, g ik is the species contribution of the kth source to the ith sample, f ki is the jth species fraction from the kth source, e ij is the residual for each species, and p is the total number of independent sources.where i = 1,…, m species, j = 1,…, n samples, k = 1,…, p sources.
In relation to the input for the PMF model, it is not necessary to utilize all of the collected data.This is because the PMF model operates under the fundamental assumption of non-reactivity and mass conservation [45].Without prior knowledge of the sources of VOCs, the technique generates source profiles and time series matrices that depict the contribution of each source to the overall VOCs emissions.In the PMF, it is a requirement that the sources possess species values that are non-negative.In addition, it is not permissible for any sample to have a negative contribution originating from a source.
To solve the PMF equation, enabling the examination of the distribution for each species and evaluation of the solution, stability, an objective function can be minimized by considering the inherent uncertainties associated with each observation, as indicated in the following equation: where sij represents the uncertainty estimate of a data point corresponding to the jth species measured in the ith sample.The analysis was conducted using software version 5.0, and Eqs. 3 and 4 were utilized to compute the uncertainties associated with the input data sets.When the concentration of VOCs falls below the MDL, Eq. 3 is applied; conversely, Eq. 4 is utilized for VOCs concentrations that exceed the MDL.However, in cases where values fell below the detection limit, they were replaced with half of the detection limit.The overall uncertainty for these substituted values was determined to be 5/6 of the detection limit, as indicated in the following equation: where MDL represents the detection limit, and the error fraction can be adjusted within the range of 5-20% depending on the concentration level [46].In this study, the Error fraction was determined through empirical (1) means, specifically set at 10% [47], based on their previous experience.
The selection of appropriate species for the PMF analysis was determined by considering the following: (1) the selected VOCs species have the highest photochemical reactivity; (2) they serve as the fundamental tracer for the target industrial byproduct around the study area; (3) they were detected in significant amounts; and (4) highly reactive species with rapid atmospheric reactivity were also omitted.Finally, species that serve as effective indicators for identifying pollution sources were appropriately selected and used for the analysis.

Performance verification of SPI-TOFMS
The integrated SPI-TOFMS instrument was characterized using a standard mixture comprising typical 48 VOCs from the Photochemical Assessment Monitoring Station (PAMS-48), as presented in Additional file 1: Table S1. Figure 4 illustrates the linearity and sensitivity parameters of benzene standard solutions within a concentration range of 1, 5, 10, 20, and 50 ppbv.The performance curves demonstrated a high level of linearity, with an R 2 value of 0.99.This indicates that the SPI-TOFMS technique demonstrates excellent sensitivity and repeatability for VOCs detection.The determination of the MDL and MQL was conducted through the measurement of the signal-to-noise ratio.MDL was defined as an S/N ratio of 3, while MQL was determined with an S/N ratio of 10.The MDL and MQL were determined to be 0.090 ppbv and 0.300 ppbv, respectively.The MDL presents the remarkable sensitivity of the technique, with a relative standard deviation (RSD, n = 4) of 4%.Moreover, the mass resolution surpassed a value of 1500.This suggest that the SPI-TOFMS instrument demonstrates analytical reliability and stability when used for the detection and monitoring of VOCs and other atmospheric pollutants in complex atmospheric environments, offering acceptable precision and exhibiting a high level of sensitivity.

Analysis of the navigation monitoring data
The VOCs commonly found in the atmospheric environment were identified and listed in Table 1 and Additional file 1: Table S2.A total of 41 VOCs constitute 99.8% of the total mass of the detected VOCs species during multiple series of navigation monitoring experiments conducted throughout the experimental period.These VOCs serves as a typical tracer for different sources of environmental emissions.The VOCs that are commonly found include n-pentane, styrene, hexane, and toluene.Alkanes and aromatics constituted the primary VOCs species that were assessed during the field detection experiments conducted for monitoring purposes.

Source profile identification and PMF analysis
In the present study, the PMF technique was employed to analyze the measured VOCs at various monitoring sites, leading to the identification of seven distinct factors.These factors profiles are presented in Fig. 5 and Additional file 1: Table S3.The major group of VOCs detected  during the monitoring experiment were alkane, aromatics, and alkenes.Alkanes were found to be the most prevalent group of compounds detected, with their primary sources being vehicle exhaust and industrial processes.This makes them a significant source of VOCs in urban atmospheres [48,49].Factor 1 reveals a high abundance of n-butane, ethane, propane, and i/n-pentane, along with smaller amounts of heptane, hexane, and methyl chloride.n-butane and ethane were identified as dominant VOCs species, collectively constituting 60% of the overall VOCs within the source profile.Ethane, propane, and butane are natural gases [50], constituting 20% of the total VOCs detected in this source profile.These serve as tracers for vehicle emissions, specifically those originating from diesel combustion engines.Based on the PMF analysis, Factor 1 has been identified as diesel vehicular emissions.In Factor 2, the presence of styrene accounts for 70% of the total identified VOCs species in the profile.In addition, a significant amount of butane and ethane compounds were also identified in the profile.Styrene and butane, has been identified as fundamental tracers for industrial solvents, as stated by Gabriel [51].Furthermore, Borbon et al. [52] assert that styrene, butane, and ethane are frequently employed as solvents in the painting, printing, and textile industries.Hence, Factor 2 is associated with paint and varnish, primarily for coating and ink applications in the rubber painting, printing, and textile industries in Suzhou [53,54].Factor 3 profile suggests ethane as the highest contributor to the VOCs accumulation, constituting approximately 80% of the total VOCs.In addition, compounds such as butane, heptane, and octane were identified in this profile, collectively accounting for 15% of the total VOCs species present.For a more precise elucidation of this source, it should be noted that these VOCs are primarily emitted from biomass burning, specifically biofuel, as well as certain vehicular emissions, such as those from diesel combustion engines.Vehicular emissions are linked to a certain proportion of other VOCs species, such as pentane, benzene, trimethylbenzene, and xylene, which were not detected in this particular factor.Furthermore, the significant presence of ethane suggests that Factor 3 has been identified as biomass/biofuel burning [55].
Factor 4 is predominantly characterized by the presence of trimethylbenzene, which constitutes 65% of the total VOCs species in this source profile.In addition, significant amounts of xylene, benzene, and styrene are also detected.These compounds are frequently released as by-products of industrial processes that involve the use of organic solvents, such as painting, printing, and coating [52].With regard to factor 5, the highest proportions of butane, ethene, methyl chloride, and xylene were observed.These species have a close association with vehicular emissions, particularly those resulting from diesel engine combustion.This process releases significant amounts of butane, ethane, and xylene [56].Thus, Factor 5 was refer to as a diesel engine vehicular emission.Factor 6 is characterized by a significant presence of lower alkanes, such as ethane, pentane, and heptane, along with notable concentrations of benzene, styrene, and xylene.The presence of a significant proportion of lower alkanes and some certain aromatic compounds has been identified as the reliable tracer of emissions from vehicles powered by gasoline [57].Therefore, Factor 6 is associated with gasoline evaporation.The seventh factor profile, which is factor 7, is predominantly influenced by a significant quantity of octane.Octane is frequently utilized as a solvent in various industries such as paint, adhesives, and liquid process photocopiers [57].

VOCs source contribution and fingerprint in Suzhou compare with other cities
Over the past three decades, Suzhou has undergone rapid economic growth, urbanization, and industrialization, resulting in a marked escalation in vehicular traffic.According to the PMF analysis results, the primary emission sources contributing to the total VOCs and predominant anthropogenic activities in the city are vehicular emissions; industrial activities such as coating, painting, and printing; and biofuel burning.These sources are similar to those observed in Nanjing city [58].Figure 6 shows the percentage contribution of the source profile of anthropogenic VOCs in the city.Li et al. [59] conducted a comparison of total VOCs emissions and their concentration levels in Suzhou with those of major cities in China, such as Beijing, Tianjin, and Shanghai.As an industrial hub of Jiangsu Province, Suzhou has the most concentrated industrial enterprises and the densest population with the highest car population, making it the only city in Jiangsu Province with a population of more than 12 million [60].Among the various sources of atmospheric VOCs in the city, vehicular emissions play a significant role.Specifically, vehicular diesel emissions contribute to 27% of total VOCs emissions, while gasoline evaporation accounts for 15% of the total VOCs emissions.Considering gasoline evaporation and diesel vehicular emissions as part of vehicular emissions, this is marked as the primary source of atmospheric VOCs in the city, comprising over 40% of the total ambient VOCs detected by the SPI-TOFMS and categorized using the PMF model.Industrial processes and solvent usage involve the use of organic solvents by industries and daily life activities (including coating, painting, and printing, while daily life activities encompass building, decoration, and cleaning) in daily production and manufacturing processes, leading to the large-scale production of various varieties of VOCs into the atmosphere.During our navigation experiment, SPI-TOFMS was able to detect various VOCs related to dyeing, printing, coating, and painting industries (including styrene, butane, ethane, trimethylbenzene, xylene, and benzene).These VOCs are responsible for 36% of the total VOCs detected by the instrument, thereby establishing them as the second most significant source as determined by the model.Biofuel burning was identified as the third major source of VOCs emissions in the city, following vehicular emissions and industrial solvents.contributing 22% to the total detected ambient VOCs.Therefore, vehicular emissions can be considered a significant contributor to VOCs in the city.
In general, the concentration of VOCs follows the order of roadside, industrial sites, and residential areas, which can be attributed to the increasing number of vehicles in Suzhou [61].Several studies conducted in various Chinese cities consistently show that motor vehicles are the main contributors to ambient VOCs [62,63].For instance, research conducted in Shanghai has confirmed that motor vehicles play a substantial role in VOCs emissions [63].Considering the predominant industrial activities in Suzhou, which encompass painting, coating, dyeing, and textile production, it can be deduced that the principal emissions arising from these sectors are organic solvents employed in their routine operations.

VOCs profiling in wastewater treatment
VOCs are the primary constituents of the odor emitted from wastewater and sewage sludge composting plants, and they are generally associated with odorous nuisances and health risks.In our navigation experiment around the wastewater treatment plant, we selected four major treatment units of the composting plant as monitoring locations, namely, the dewatered unit, dewatered sludge, fermentation workshop, and product units.The product units exhibited the highest concentration of VOCs emissions at 13.367 ppbv, followed by the fermentation workshop and dewatered sludge, with emissions of 9.290 ppbv and 3.240 ppbv, respectively, as presented in Table S4 in the supporting information.The detected VOCs families included halogenated hydrocarbons, aldehydes, and aromatic compounds.Halogenated hydrocarbons, particularly bromodichloromethane, were the most abundant compounds in all samples detected by SPI-TOFMS in this area.The product units differed from the other sampling units, as their typical compounds were toluene, benzene, and styrene, all of which were the products of the latter stages of composting.Among the treatment units, the fermentation workshop units had the highest halogenated hydrocarbon (bromodichloromethane) concentration.Acrolein, bromodichloromethane, and benzene were the major contributors, contributing to the increased atmospheric VOCs and ozone formation potential in the surrounding air.

VOCs tag identification library
Pollution leakage accidents can occur abruptly.To enhance the efficacy of pollution control techniques, it is imperative to promptly identify the distinct characteristics of VOCs that contribute to pollution and trace their emission sources.Therefore, it is imperative to establish a comprehensive database library that includes a representative VOCs tracer for each emission source.This library will serve as a valuable resource for efficiently tracking different sources of emissions during the control process.
Additional file 1: Table S5 provides a comprehensive comparison between our navigation experimental SPI-TOFMS and traditional techniques, specifically GC-MS, portable GC-FID, and GC-MS.The comparison spans various critical aspects, including the detection of specific VOCs, sorbent phase, qualitative/quantitative mode, sample types, and overall performance.Our navigation experimental setup, enabled by SPI-TOFMS, facilitates real-time and on-site monitoring, capturing the dynamic nature of VOC emissions in diverse environmental settings.This stands in contrast to traditional techniques that may provide delayed or less flexible measurements.
Table 2 displays VOCs tracers used to identify different sources.Acrolein, bromodichloromethane, and benzene serve as tracers for sewage treatment, while toluene and heptane are utilized as tracers for printing and dyeing.Butane, octane, and xylene are employed as tracers for oil-spraying plants.These results are consistent with previous studies, which have shown that toluene is the predominant pollutant from the printing and dyeing industries [52][53][54].Similar to the findings of Ravina et al. [64], it was observed that benzene and bromodichloromethane serve as the primary pollutants in sewage treatment plants.These findings demonstrate the stability and reliability of our instrument for emergency monitoring.
According to the real-time navigation monitoring results of SPI-TOFMS in Suzhou, a wide range of VOCs were detected.Furthermore, several emission sources, including wastewater and sewage sludge, printing and dyeing plants, and oil spraying plants, were identified and analyzed using VOCs tracers associated with these sources.Attributes characterizing the VOCs species are shown in Table 2. SPI-TOFMS is a promising technique for real-time detection at gas leakage accident sites, providing an accurate reference for emergency detection and monitoring VOCs emission sources.

Ozone formation potential (OFP)
The ozone formation potential (OFP) serves as a significant index for quantifying and elucidating the role and contribution of individual VOCs in the formation of ozone in the atmosphere.Theoretically, controlling the NO x emission and VOCs in the atmosphere will lead to a reduction in atmospheric ozone.The relationship between ozone formation and precursor concentration is not solely characterized by a linear correlation.The quantity of ozone in NO x conditions is contingent upon the level of radicals produced from the reaction of VOCs.According to Carter et al. [65], the formation of ozone is limited at lower NO x concentrations lower.Previous studies have proposed various approaches to investigating the role of specific VOCs in the photochemical process [66,67].In this study, the concept of maximum incremental reactivity (MIR) is used to evaluate the ozone formation potential of the individual VOCs.The techniques and process employed to establish the MIR scale, as outlined in this study, are thoroughly described by Carter et al. [68], and references therein, and those reports should be consulted for details.Briefly, the reactivity scale is determined by calculating the relative ozone impacts, expressed as mass of additional ozone formed per mass of VOCs added to the emissions, for various compounds under various atmospheric conditions, given a chemical mechanism for the compounds and other relevant atmospheric species, models for various atmospheric conditions, and a modeling and reactivity assessment procedure.Incremental reactivity can be conceptualized as the result of two underlying factors: (1) kinetic reactivity, which refers to the proportion of emitted VOCs that undergo chemical reactions in the specific pollution scenario under consideration and (2) mechanistic reactivity, which quantifies the amount of ozone formed in relation to the amount of VOCs that have reacted.The significance of this concept is that this provides a means to factor out to a certain extent, the impact of a VOC's reaction rate from all the other mechanistic aspects which affect reactivity.Since the only aspect of the VOC's mechanism which affects kinetic reactivities is the VOC's rate constants, which generally are known, the ability of a mechanism to predict kinetic reactivities are not considered to be particularly uncertain.MIR computes the OFP of individual VOCs by considering its reactivity mechanism, taking into account the impact of the different reaction mechanisms and VOCs and NO x ratio, on the formation of ozone [24].The ozone formation is determined by the multiplication of the concentration of each VOCs species by its corresponding MIR coefficient as expressed by the equation: where VOCs (i) represents the concentration of individual VOCs, and MIR (i) denotes the maximum incremental reactivity of VOCs species i.
The results presented in Table 1 highlight the primary factors that contribute to ozone formation in Suzhou.Ethane, trimethylbenzene, xylene, toluene, and styrene have been identified as the predominant substances contributing to the formation of ozone in the city, collectively contribute for 67% of the total ozone production.Despite alkanes having the highest volume fraction among the total identified VOCs, their photochemical reactivity is comparatively low, leading to a diminished impact on ozone formation.On the contrary, aroma compounds, despite their lower percentage contribution to the total VOCs in comparison with alkanes, play a more significant to the formation of ozone in Suzhou.

Dynamics analysis of NO x and O 3 in Suzhou
The average concentration of NO x fluctuated between 6.540 and 21.100 ppbv during winter and 3.950 to 16.870 ppbv during summer, respectively.The trends of NO x concentrations in Suzhou were similar to those in Shanghai [69].As illustrated in Fig. 7, the daily average concentration of NO x in Suzhou exhibited a bimodal OFP = VOCs (i) × MIR (i) distribution: the first peak appeared between 08:15 and 08:45 a.m. in the morning, and the second peak occurred between 5:45 and 6:15 p.m. in the evening.Interestingly, the concentration of the second peak was significantly lower than the morning peak, possibly linked to higher emissions from diesel vehicles [70].High concentrations were observed during the winter morning rush hours, while lower concentrations were recorded during summer nights.The seasonal patterns of NO x in our study align with those reported in other studies [71,72].In general, NO x concentrations were higher in winter, and the change in the concentration trend was associated with meteorological conditions.The analysis suggests that NO x concentrations in the city are lower during summer than the winter period due to temperature inversions.In contrast, O 3 concentrations are higher during summer than winter due to longer days and higher solar radiation, leading to a higher rate of photochemical reaction.In addition, during summer, the reduction of NO x leads to an enhancement in OH radicals, resulting in higher O 3 production.This enhancement of summer O 3 is offset by a decrease in VOCs.

Conclusion
The emissions sources of volatile organic compounds (VOCs) were traced and identified in China's city Suzhou.The analysis of their contribution to ozone formation was conducted using a single-photon ionization-time-of-flight mass spectrometer (SPI-TOFMS).The study utilized the positive matrix factorization (PMF) model to conduct an analysis of the apportionment of VOCs sources.The results of the analysis identified three main sources of VOCs emissions: vehicular emissions (F1, F5, and F6), industrial solvents (F2, F4, and F7), and biofuel burning (F3).Alkanes were the most prevalent VOCs species, constituting 60% of the overall VOCs.Aromatic hydrocarbons and alkenes were also present, albeit in lower quantities.Aromatics and alkenes were observed to make a substantial contribution to ozone formation, as evidenced by the results of the maximum incremental reactivity coefficient (MIR).The study suggests that efficient control of ozone formation in Suzhou can be achieved by reducing emissions of ethene, trimethylbenzene, xylene, styrene, and toluene originating from solvent-based industries, vehicles, and biofuel combustion.The primary contributors to the potential for ozone formation in Suzhou were identified as sources related to transportation and industrial solvents.In addition, the utilization of SPI-TOFMS facilitated the development of an identification library for emergency monitoring in the event of pollution accidents, thereby enabling the precise localization of emission sources.The study proposes that the use of SPI-TOFMS enables the rapid and direct monitoring and tracing of VOCs emission sources in the atmospheric environment.This method shows promise for efficiently inspecting and controlling air pollution.Furthermore, the development of targeted pollution control strategies requires collaboration between researchers, policymakers, and industries.Implementing effective emission reduction measures, promoting sustainable practices, and adopting cleaner technologies are pivotal in mitigating ozone pollution.

Fig. 1
Fig. 1 Schematic diagram and photograph of the vehicle

Fig. 2
Fig. 2 Map of the study area: Suzhou China

Fig. 5
Fig. 5 Factor profiles derived from the PMF models used to predict the VOCs sources in the Suzhou industrial park

Fig. 6
Fig. 6 Source contributions to ambient VOCs in Suzhou

Fig. 7
Fig. 7 Daily average concentration of nitrogen oxide (NO x ) in Suzhou AbbreviationsCI Capillary inlet ICP Inductively coupled plasma IP Ionization potential LOD Limits of detection LOQ Limits of quantification MI Membrane inlet MIMS Membrane inlet mass spectrometry MIR Maximum incremental reactivity MoBiMS Modular biological mass spectrometry MDL Method detection limit MQL Method quantification limit MS Mass spectrometer OBM Observation-based model OFP Ozone formation potential O 3 Ozone PDMS Polydimethylsiloxane PMF Positive matrix factorization PTRMS Proton transfer reaction mass spectrometry RSD Relative standard deviation SPI Single-photon ionization TOF Time-of-flight

Table 1
Measured concentrations selected for PMF model simulation [68]R denotes maximum incremental reactivity, and updated MIR values are obtained from[68]b [VOC] × MIR, ppbv

Table 2
Emissions source and measured VOC at the navigation monitoring experiment site in SuzhouBold values signify the presence of higher concentration compared with others or signify the major tracer