Monitoring Volatile Organic Compounds in Air Using Passive Sampling: Regional Cross-Border Study between N. Macedonia and Kosovo

The aim of the study was to monitor volatile organic compounds (VOCs) in outdoor air in nine cities/towns across the Republic of N. Macedonia and the Republic of Kosovo for nine consecutive months, using the Radiello ® passive and diffusive samplers, and gas chromatography-mass spectrometry (GC-MS). In parallel, employing static headspace gas chromatography-mass spectrometry (SHS-GC-MS) technique, VOCs from different types of commercially available gasoline and diesel fuels in the monitored cities were analyzed in order to obtain the chemical profiles and to evaluate the presence of their components in ambient air. GC-MS analyses indicate that the gasoline fuel components (BTEX: benzene, toluene, ethylbenzene, and xylenes) and diesel fuel components ( n -undecane, n -dodecane, n -tridecane, n -tetradecane, and other linear alkanes) account for approximately 60% of VOCs in the outdoor air in the sampling sites. The bulk of the VOCs in the all-sampling locations regardless of the season come from fossil fuels (automotive fuels) and only minor fraction comes from biogenic emission which is dominated by monoterpenes. Furthermore, the ratios of benzene to toluene (B/T) and xylenes to ethylbenzene (X/E) were used to assess the origins of VOCs emissions in the outdoor urban air at the monitored sites.


INTRODUCTION
In this day and age, one of the main global challenges is the exposure of human beings to air pollution (Leung, 2015;McDonald et al., 2018;Seinfeld, 2004).According to the World Health Organization (WHO), air pollution contributes to over seven million annual deaths worldwide, accounting for approximately one-quarter of all fatalities (Wohlgemuth et al., 2020;WHO, 2004).In recent years, growing awareness about air quality has raised concerns regarding hazardous air pollutant (HAP) emissions on a global scale, prompting many states to develop strategies aimed at improving outdoor air quality (Gee and Sollars, 1998;Mohamed et al., 2002).
Using the information from the US Environmental Protection Agency (EPA) and the European Environment Agency (EEA), a diverse group of VOCs, which includes BTEX (benzene, toluene, ethylbenzene, and xylenes), have been classified as hazardous air pollutants (HAPs) (Huang et al., 2018;U.S. EPA, 2022).Even though BTEX compounds share similar chemical structures, their physicochemical properties differ (Słomińska et al., 2014).In Fig. S1, presented in the supplementary material, literature data on the lifetimes, and total decomposition times of BTEX compounds in atmospheric air presented (Słomińska et al., 2014).It is essential to pinpoint the sources of anthropogenic VOCs to effectively alleviate air pollution and improve human health.
VOCs are emitted into the atmosphere from a range of sources, including anthropogenic activities like industrial processes, transportation, petroleum refineries, domestic processes, fossil fuel-burning power plants, textile cleaning, construction activities, and pharmaceutical industries, as well as from biogenic or natural sources like vegetation, volcanoes, anaerobic moors processes, and natural forest fires (David and Niculescu, 2021;Gu et al., 2021;Huang et al., 2018;Mohamed et al., 2002).Biogenic volatile organic compounds (BVOCs) account for around 80%-90% of all VOCs emissions into the environment annually (Antonelli et al., 2020;Laothawornkitkul et al., 2009).
In urban setting significant contributor to VOCs is the transportation sector, which heavily relies on petrochemical fuels (gasoline, diesel, and fuel oils).In recent years, in the western countries the automotive VOC emissions have steadily dropped due to improved vehicle (gasoline and diesel) engines and fuel technology (McDonald et al., 2018).The same authors have concluded that in urban air, VOCs arising from transportation have decreased and the organic compounds from chemical products (so-called VCP) have drastically increased.In other words, in urban settings there is a shift from transportation-related sources toward VCPs (McDonald et al., 2018).
In 2019, a World Bank study revealed that residents of the Balkans and Eastern Europe experience elevated exposure to fine particulate matter (PM2.5 and PM10) in comparison to their Western European counterparts, with the Balkans hosting seven of Europe's top ten most polluting coal-fired power plants of which one is located in the vicinity of the capital of Kosovo (Prishtina) (The World Bank, 2019a, 2019b, 2019c).Regional studies are needed to make an to locate the sources of anthropogenic VOCs to effectively alleviate air pollution.
It is important to note that here are no operating refineries in North Macedonia and Kosovo.From 1982 until 2013 (with some breaks), the OKTA (refinery in Macedonia) near the capital city of Skopje, produced fuels according to the European standards (EU, 2007).Since January 2013, the refinery is inactive, but the storage capacities are used for imported petroleum derivatives which must satisfy the requirements of the European Union.The other large chemical factory OHIS in Skopje shut down the production about 3 decades ago and is not expected to contribute as emitter of VOCs.
It is crucial to have preliminary data on the composition of VOCs and to select the proper air monitoring methods, as there is currently no available data about their presence and content in the Western Balkan countries, especially Kosovo and Macedonia.The study presented herein, was designed for monitoring volatile organic compounds (VOCs) in the atmospheric air at nine cities/towns near the North Macedonia-Kosovo border over several months.Passive/diffusive samplers were used, followed by gas chromatography-mass spectrometric (GC-MS) analyses.The results were compared to VOC profiles found in gasoline and diesel fuel components to identify emission sources at the monitored sites.

Sampling Sites
The monitoring sites were strategically selected in order to represent different urban and suburban settings within the chosen cities/towns along the Kosovo and North Macedonia borders.The selection criteria included several factors: avoiding direct proximity to emission sources, ensuring optimal air circulation, avoiding obstacles or barriers, and ensuring both accessibility and security.The sampling sites were positioned in close proximity to major traffic routes, industrial zones, and residential regions.The samples were collected from nine locations (Fig. 1) with the following characteristics and differences: 1. Tetovo (41.9978°N; 20.9626°E) is a city in North Macedonia, surrounded by the Sharr and Dry Mountains, and has a population of 84,770 people (Wikipedia, 2021).The sample location was one kilometer from the main road, in an open area surrounded by residential buildings (464 m a.s.l.).

Sampling Methods and Sample Preparation
Radiello® passive/diffusive samplers, designed and developed by the Fondazione Salvatore Maugeri in Padova, Italy (Radiello®, 2023), were placed in nine sampling locations (Fig. 1) to monitor VOCs in the atmospheric air of the border region between the Republic of North Macedonia (Skopje, Tetovo) and the Republic of Kosovo (Prishtina, New Prishtina, Prizren, Dragash, Rahovec, Obiliq, and Elez Han).
The Radiello® passive sampler (RAD 130) was selected for its ability to effectively capture a wide range of VOCs in outdoor air, its ease of use, low maintenance requirements, and compatibility with our research objectives.Additionally, it operates without the need for pumps or electricity, further enhancing its suitability for our study (Radiello®, 2023).
The adsorbing cartridge is coaxially housed within the white diffusive body, allowing the sampling of BTXE, as well as other VOCs, by diffusion through the microporous polyethylene membrane and adsorption onto the activated charcoal cylindrical cartridge (Radiello®, 2023).
Each RAD 130 sampler was pre-conditioned and calibrated according to the manufacturer's guidelines before application.The samplers were placed at an approximate height of 3 meters above ground level to assess pedestrians' breathing zone exposure.Additionally, they were protected from rain and bright sunshine using mountable polypropylene shelters (Cat.No. RAD196).
Air samples were collected from each site monthly, spanning from April to December 2022.A total of 9 air samples were collected at each monitoring site throughout the study, culminating in a total of 81 samples.A new cartridge was used for each sampling period, with a sampling duration of 28 days.This timeframe was chosen to capture seasonal variations and potential short-term events that might affect air quality.The exposed cartridges were securely stored in sealed plastic bags at a temperature of 4°C until the GC-MS analysis.
The charcoal cartridges used during sampling were transferred into vials and fortified with 100 µL of anisole (internal standard), followed by 2 mL of a dichloromethane.The vials were then sealed with septum caps.The samples were gently shaken for 30 minutes at ambient temperature, and 1000 µL of the liquid phase was transferred to a vial for GC-MS analysis.
The various types of gasoline and diesel fuel from different gas stations were collected in April 2022 at the monitoring sites.The samples were instantly refrigerated at 4°C at the sampling locations before being transferred to the laboratory refrigerator.

GC-MS Analysis
The chromatographic separation of VOCs was carried out using an Agilent 6890N gas chromatograph coupled to a single quadrupole 5975B mass selective detector.HP-5MS (30 m × 0.25 mm, 2.25 µm, Agilent Technologies Inc.) stationary phase was used with helium as a carrier with a constant flow of 1.0 mL min -1 .The GC was initially set at 35°C for a duration of 5 minutes and then gradually raised to 90°C at a rate of 5°C min -1 for 3 minutes, followed by another increase to 280°C.The injector temperature was set at 240°C and 2 µL of the sample was injected using the split-less mode.The MS source was set at 230°C, the quadrupole at 150°C and the m/z scan range of 35-500 amu.The total run time per analysis was 40 minutes.
For the analysis of gasoline and diesel fuels, an Agilent 7697A Headspace autosampler was used.A 1.0 µL sample was placed in the 22 mL Headspace injection vial, secured with septum and heated at 145°C for 15 minutes.The temperature of the loop was 150°C, with the fill pressure of 15 psi and fill time of 0.25 min.The transfer line was set at 160°C.Upon the completion of the process the sample was introduced to the GC-MS with the same method as described for the liquid injection.
The analysis of collected samples of ambient air, gasoline, and diesel fuels was done using the Agilent MassHunter v.10.0 software or ChemStation.Identification of compounds was accomplished by comparing their retention times with available standards, and also by using mass spectra and matching with NIST Library of mass spectra (NIST MS 2011).The compounds for which standards were not available were identified using the NIST Mass Spectral Database with probability values greater than 85%.
Mass percentages of individual VOCs in ambient air were calculated by dividing each compound's surface area by the response factor of anisole.The resulting values were then divided by the total mass of identified VOCs and multiplied by 100%.
To determine the mass percentage for each group (Alkane C4-C9, Alkane C10-C24, BTEX, C9-C11 aromatics, oxygenated volatile organic compounds (OVOCs), and natural volatile organic compounds (NVOCs)), mass percentages of compounds within each group were summed.This method allowed for precise assessment of VOC distribution in the ambient air, providing insights into air quality and composition.

RESULTS AND DISCUSSION
Based on the HS-GC-MS analyses of the commercially available automotive fuels in Kosovo and Macedonia it was evident that they have very similar profiles.The analyses indicate that the dominant gasoline fuel components are toluene, ethylbenzene, and xylenes and diesel fuels are dominated by alkanes (n-undecane, n-dodecane, n-tridecane, n-tetradecane, and other linear alkanes).These compounds can also be detected in the collected samples from the above-mentioned locations.
In Fig. 2 the mass percentages (%) of the identified organic compounds in ambient air, gasoline, and diesel fuels from April 2022 are depicted.The data indicates that the primary VOCs in air samples from monitoring sites in April 2022 are BTEX, C9-C11 aromatics, and alkane C10-C24 species, which collectively account for approximately 84% of the total compounds.These compounds closely resemble those found in different types of gasoline and diesel fuels used in North Macedonia and Kosovo.
In Fig. 3, the average seasonal mass percentages (%) of all categories of VOCs detected at the monitoring sites in 2022 are illustrated.
The BTEX, C9-C11 aromatics, and alkane C10-C24 species are the main VOCs detected at monitoring sites, accounting for about 70% of total compounds, and they are most likely from road transport vehicles (Kerchich and Kerbachi, 2012).The profiles of all the samplers are very similar, with exception of those in Dragash and Rahovec.However, in these two locations, alkanes C10-C24 (diesel components: n-tridecane, n-tetradecane, n-pentadecane, and n-hexadecane) dominate with a  mass distribution of about 25%.This may be because Dragash and Rahovec are agricultural areas that use a higher proportion of heavy tractors and diesel-powered equipment than other monitoring sites.
The outdoor air in Obiliq and Elez Han has higher levels of oxygenated volatile organic compounds (OVOCs) than those in other sites, with 34% in Obiliq and 26% in Elez Han.This is because of the use of coal and oxygenated solvents by the industrial activity in these areas, such as Power Plants A and B in Obiliq and the cement and plastic factories in Elez Han.Plant emissions are the main natural source of OVOCs, but their levels increase in all samples during summer due to the photochemical oxidation of VOCs in the environment (Mellouki et al., 2015).
Biogenic volatile organic compounds (BVOCs), such as terpenes, are emitted from the leaf surface of chlorophyll-containing plants during photosynthesis (Miyama et al., 2020).The average mass percentage of BVOCs emitted varies with the temperature of the atmosphere (Miyama et al., 2020).Based on the data, one can conclude that the quantity of BVOCs in the atmosphere increases as the temperature of the atmosphere rises (Antonelli et al., 2020;Miyama et al., 2020).As a result, the mass percentage of isoprene, monoterpenes, and sesquiterpenes, which are the main components of BVOCs, in the atmosphere increased in the spring and summer compared to autumn and winter due to higher temperature, light intensity, and vegetation activity (Antonelli et al., 2020;Miyama et al., 2020).Among all monitoring sites, Dragash had the highest quantity of BVOCs (10.65%) compared to other sample locations.It was located in the Sharr Mountains, with abundant coniferous trees such as pine and other green vegetation.
Tetovo and Prishtina had the highest average distribution of mass percentages of BTEX in atmospheric air, with 50% and 49%, respectively.These cities are likely to have high levels of traffic emissions, industrial activities, and gasoline stations, which are the main sources of the VOCs.
With the increasing number of passenger cars, complex issues appear, with serious effects on emissions and fuel use (Kostenidou et al., 2021;Lopes et al., 2014;Saliba et al., 2017;Zhu et al., 2016).Biofuel production and use have expanded dramatically in recent years to lower greenhouse gas emissions and support renewable energy sources (Karavalakis et al., 2009;Lopes et al., 2014;Rounce et al., 2012;Shi et al., 2006).One of the most significant distinctions between the two types of diesels is that biodiesel is more viscous and contains approximately 10-12% oxygen, whereas standard diesel has no oxygen (Karavalakis et al., 2009;Lopes et al., 2014).Furthermore, the elevated oxygen content in biodiesel fuel enables it to burn more completely, resulting in a reduction of VOC emissions, such as benzene, toluene, and octane, by 60-80% in comparison to regular diesel (Karavalakis et al., 2009;Lopes et al., 2014).Biodiesel is both biodegradable and non-toxic (Lopes et al., 2014).Thus, regular diesel is more hazardous to human health compared to biodiesel blends, and higher biodiesel percentages result in lower total VOC emissions (Lopes et al., 2014).However, the impact of biodiesel on VOC emissions can vary significantly depending on the vehicle, engine technology, and driving patterns (Lopes et al., 2014).Total hydrocarbon (THC) emissions were found to be higher in ethanol and diesel fuel mixes, while biodiesel blends produced lower THC emissions than pure diesel fuel (Shi et al., 2006).Research suggests that a combination of biodiesel, ethanol and diesel fuel can be a feasible alternative fuel option for diesel engines (Karavalakis et al., 2009;Rounce et al., 2012;Shi et al., 2006).
The ratios of benzene to toluene (B/T) and xylenes to ethylbenzene (X/E) serve as significant indicators used to assess the origins of volatile organic compounds in the outdoor urban air (Abtahi et al., 2018;Caselli et al., 2010).Based on the previous studies, it has been proposed that B/T ratios within the range of 0.23-0.66indicate that traffic vehicles constitute the main contributors to VOC emissions in the atmosphere.A B/T ratio lower than 0.23 suggests that volatile organic compounds emanate predominantly from stationary sources, while a B/T exceeding 0.66 indicated that VOC emissions primarily originate from both stationary and mobile sources in the ambient air (Abtahi et al., 2018).Furthermore, literature research suggests that X/E ratios within the range of 2.8-4.6,indicate vehicular traffic as the predominant source of VOCs in the ambient air (Caselli et al., 2010).Ratios of the B/T and X/E calculated from data obtained locations and other areas obtained from the literature are presented in Table 1.
Compared with our findings, the B/T ratios ambient air determined in Orleans (France) and Tehran (Iran) are higher than the B/T ratio in this study, which ranged from 0.61-1.21.The values  Abtahi et al., 2018) obtained in London (B/T = 0.65) are comparable to the values in Tetovo (0.62), Skopje (0.61), and Prizren (0.66).The X/E ratios in our study were from 1.7-2.72 and they were below the suggested range of 2.8-4.6.The average B/T and X/E ratios observed in our work indicate that VOC emissions in the ambient air mainly come from stationary and mobile sources.All in all, if one takes into account the chemical profiles of commercially available gasolines and diesel fuels, they are very similar in Macedonia and Kosovo.The bulk of the VOCs in all of the sampling locations, regardless of the season, come from fossil fuels (automotive fuels) and only minor fraction (less than 10%) comes from biogenic emission.The highest content of BVOCs was observed in the mountain town of Dragash, where in the summer time they were above 10%.The biogenic emissions were dominated by the monoterpenes limonene and α-pinene.All of the above, points toward the need for a more detailed study in the speciation of AVOCs using complementary sampling techniques, which will help us locate the key emission sources.

CONCLUSIONS
Volatile organic compounds (VOCs) have not been monitored, in most of the former Yugoslav countries, including the N. Macedonia and Kosovo.Herein, the Radiello® diffusive passive sampler devices and subsequent GC-MS analysis were applied to monitor VOCs in nine cities/towns near N. Macedonia and Kosovo border.Utilizing HS-GC-MS method we were able to make a profile of VOCs of different types of commercially available gasoline and diesel fuels in the monitored cities and evaluated the presence of their components in ambient air.
The rank order of organic groups of VOCs based on their mass percentages (%) emission in the outdoor air is as follows: alkane C10-C24 > BTEX > C9-C11 aromatics > OVOCs > NVOCs > alkane C4-C9 > PAHs > halocarbon > alkenes C4-C7.The ratios of benzene to toluene (B/T) and xylenes to ethylbenzene (X/E) point that the primary sources of VOC in the ambient air of nine cities

Fig. 2 .
Fig. 2. Comparison of volatile organic compounds (VOCs) in commercial gasoline, commercial automotive diesel, and in outdoor air (April 2022) as determined by GC-MS.

Table 1 .
Comparison of B/T (Benzene/Toluene) and X/E (Xylenes/Ethylbenzene) ratios in sampled locations monitored and other areas based on the literature researches.