Occurrence and Distribution of Persistent Organic Pollutants (POPs) in the Atmosphere of the Andean City of Medellin, Colombia.

Passive air sampling (PAS) was used to evaluate organochlorine pesticides (OCPs), polychlorinated biphenyls (PCBs), polybrominated diphenyl ethers (PBDEs), polybrominated biphenyls (PBBs), hexabromocyclododecanes (HBCDDs), polychlorinated dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs), and per- and polyuoroalkyl substances (PFASs) in the atmosphere of Medellin, Colombia. This was in accordance with a special initiative of the Global Monitoring Plan (GMP), which was implemented to provide information on new and emerging persistent organic pollutants (POPs). PAS was carried out for three months (four quartiles per year) over two consecutive years (2017 and 2018), and two punctual peaks of contamination in the monitored time were observed. The rst peak was related to chlorinated compounds in the quartile 2017-Q3, in which the highest concentration of pentachlorobenzene (PeCB, 755.5 pg·m -3 ) was observed, and air concentration of PeCB in quartile 2018-Q4 (125.0 pg·m -3 ) was 3.7 times more than the initial concentration (33.7 pg·m -3 ). These results may be associated with waste incineration in or near the city. The second peak of contamination was associated with brominated compounds in the quartile 2018-Q1, which could be linked to recycling, incineration, and landll deposit of these compounds. In this peak, the highest levels of PBDEs and HBCDD-isomers were observed (Σ 10 PBDEs, 107.2 pg·m -3 and Σ 3 HBCDDs-isomers, 289.0 pg·m -3 ). With regard to the concentrations of PBDEs, HBCDDs, PCDDs, and PCDFs, these showed a slight tendency to increase between 2017 and 2018. Finally, constants concentrations of pollutants such as DDT isomers and dieldrin were observed, although these compounds were banned in Colombia many years ago. In contrast, a slight decrease in some pollutants, such as aldrin, and α-endosulfan, was observed. Our study allowed the assessment of air levels of chlorinated, brominated, and uorinated pollutants in Medellin, Colombia. These results provide an overall view of POPs levels and represent an initial attempt to identify local sources in order to monitor and surveillance the releases of these pollutants in the city and country.

Although POPs are restricted under the Stockholm Convention, they are released as by-products of chlorinated hydrocarbons, internal combustion engines, and thermal processing of waste (Cortés et al., 2016; Lee et al., 2004;Villa et al., 2016). The quantity of POPs originating from these sources is such that they are even found in quantitative levels in abiotic and biotic matrices around large Latin American and Caribbean urban areas (Bogdal et  Environmental Program, is a multilateral environmental treaty aimed at reducing and ultimately eliminating the release of these compounds into the environment and subsequent human exposure (Rauert et al., 2018b;Villa et al., 2016). One of the major pillars for the evaluation of compliance with the Convention is the monitoring of data in core media through the Global Monitoring Plan for POPs (GMP) (Magulova and Priceputu, 2016). The air is a selected matrix under the GMP, given that atmospheric transport has been identi ed as the key global dispersal mechanism for most legacy POPs (Lohmann et al., 2007;Schuster et al., 2015). Moreover, air monitoring data are important in order to identify changes in the concentrations of atmospheric contaminants over time and to provide information on their regional and global environmental transport. However, the generally low concentrations of POPs in air mean that elaborate sampling techniques, and comprehensive analytical methods and resources, are required.
Although several monitoring studies on POPs in the air at a global scale have been conducted, atmosphere monitoring of POPs remains a challenging task (Bogdal et al., 2013).  (Wania and Shunthirasingham, 2020). The use of PAS is widespread globally because of their ease of handling; low price; simple operation, as there is no need for calibration; and the fact that no electrical connection is required (Romo-Melo et al., 2018; Wania and Shunthirasingham, 2020). Passive samplers employ several different materials and methods for analysis of gases, such as polyurethane foam (PUF) disks, which have proved to be an adequate adsorbent for deposition of particles (Shoeib and Harner, 2002).
According to the literature, several studies have used passive sampling to support monitoring plans in different parts of the world. For example, the PUFs-PASs used in the present research are similar to the devices used in the regional and local monitoring of POPs in previous studies in Latin America and Colombia (Cortés et al., 2016(Cortés et al., , 2014Rauert et al., 2018aRauert et al., , 2018bSaini et al., 2020;Schuster et al., 2015).
In Colombia, passive monitoring data of PCDDs, PCDFs and dioxin-like PCBs (dl-PCBs) has been carried out in two tropical Andean cities, speci cally Manizales and Bogotá, to identify potential sources of POPs In addition, a pilot study under the GAPS network monitoring the level of PBDEs, HBCDDs and other emerging contaminants in Bogota, Colombia has been reported, the results of which showed levels of Σ 9 PBDEs of 11.7 pg·m − 3 (Saini et al., 2020).
Given the importance of urban emission sources of POPs, urban sites are a subject of extensive study due to the elevated concentration and emission of these compounds. Thus, the aim of this project was to assess the levels of OCPs, PCBs, PBDEs, PBB, PCDDs, PCDFs, HBCDDs, and PFASs in the atmosphere of Medellin, Colombia, between 2017 and 2018, using passive air sampling. This represents an important contribution to research on this topic, because PFASs levels in the atmosphere of Medellin, Colombia have not been reported, and to the best of our knowledge, this is the rst study.
Moreover, this research offers comparable information on POPs with other regions and countries to complement existing information on the distribution patterns in air of these pollutants. These results can be used to develop regulatory treaties and assess risks to human health in the city and country.

Sampling site
Passive air samplers were deployed between January 2017 and January 2019 in Medellin, Colombia. The sampling site was located at 6°15'38''N 75°34'04''W, and the average temperature at the site during the deployment period was 21.6°C. Medellin is categorized as urban, is the second-largest city in terms of population in Colombia (2,508,452 inhabitants) and is an industrialized city.
Polyurethane foam passive air samplers (PUF-PASs) were deployed in an urban site with high tra c ow.
Each sampling campaign was conducted for three months, giving four sampling quarters, referred to as quartile 1 (Q1), quartile 2 (Q2), quartile 3 (Q3), and quartile 4 (Q4), throughout each sampling year (between 2017 and 2018). The exception to this was the evaluation of PCDD/Fs and dl-PCBs, for which the four PUF samples per year were combined due to low concentrations of these contaminants in the air. PUF-PAS deployment details are listed in Table S1 in the Supporting Information (SI).
The pre-conditioned PUFs were shipped from the Laboratory of Dioxins of the Institute of Environmental Assessment and Water Research (IDAEA-CSIC) (Barcelona, Spain) to the sampling site. After collection, the samples were shipped back to the different reference laboratories for the analysis of POPs.

Volumetric air concentrations
Air concentrations (pg·m − 3 or fg·m − 3 ) were derived using the mass quanti ed in the sample divided by an effective air sampling volume (V air ). The V air was calculated using the GAPS template, version 2021_v10

Quality assurance and quality control
Air concentrations (pg·m − 3 or fg·m − 3 ) were derived from the mass of target analyte quanti ed, these values were obtained from IDAEA-CSIC (OCPs, PBDEs, PCBS, PBB, PCDDs, and PCDFs), Vrije Universiteit Amsterdam (HBCDDs), and Man-Technology-Environment research center (PFASs) divided by an V air (Table S2), these effective air sampling volume were determined using GAPS template, version 2021_v10 (Harner, 2021). The PFOA and FOSA were assumed to stay in the linear sampling phase during deployment, and air sampling volumes were calculated as the number of days the sample was deployed multiplied by a sampling rate of 4 m 3 ·day − 1 (Pozo et al., 2004a). In the cases of the analyte were not detected to convert method limits of quanti cation to pg·m − 3 or fg·m − 3 , the average air sampling volume was applied.

Results And Discussion
In general, the results showed two punctual peaks of contamination in the monitored time. The rst peak was associated with chlorinated compounds in the quartile 2017-Q3, in which the higher concentrations of PeCB and indicator PCBs were observed. Meanwhile, the second peak of contamination was associated with brominated compounds in the quartile 2018-Q1, in which the higher levels of PBDEs and HBCDD-isomers were observed, as was a slight tendency for the air concentrations of PBDEs, HBCDDs, PCDDs, and PCDFs to increase in the year 2018 with respect to 2017. Also, constant concentrations in air were observed of pollutants banned in Colombia many years ago, such as DDT isomers and dieldrin, in contrast to a slight decrease in some pollutants such as aldrin, and α-endosulfan. Finally, the results provided a general vision of the air concentrations of chlorinated, brominated, and uorinated pollutants in Medellin, Colombia, and can be a starting point for the identi cation of local sources and monitoring of the release of these pollutants in the city.

Organochlorine pesticides (OCPs) and indicators polychlorinated biphenyls (indicators PCBs)
Twenty-eight organochlorine pesticides and six indicator PCBs were monitored in the air in Medellin, Colombia during the period 2017 to 2018. The PeCB showed the highest levels of this chlorinated compounds group, generally called basic POPs, with values between 33.7 and 755.5 pg·m − 3 (Fig. 1a). These high levels of PeCB with respect to the other basic POPs was due to the contamination peak observed in the quartile 2017-Q3 (Fig. 1b).
In this period elevated emissions of this pollutant occurred, which could be associated with discharges of PCBs or waste incineration, given that the same was observed in this quartile (2017-Q3) for indicator PCBs (Fig. 1b). This association is due to the use of PeCB in dielectric uids with PCBs, and its unintentional formation in the combustion process; additionally, it is a precursor of PCBs formation under these conditions (Liu et al., 2001). The increase in air concentrations was more evident for the low chlorinated PCBs (tri and tetra-chlorinated congeners) respect to the high chlorinated PCBs, and the highest concentrations were 56.7 and 54.4 pg·m − 3 , for PCB 28, and PCB 52 respectively, in the quartile 2017-Q3 (Table S3). The air concentrations of PeCB and indicator PCBs decreased in the quartiles subsequent to 2017-Q3, which con rms the punctual contamination in 2017-Q3 of these pollutants. However, the air concentrations of PeCB in the 2018 period were higher than in 2017 and tended to stabilize at ~ 125 pg·m − 3 , which is 3.7 times higher than the initial concentration observed in 2017-Q1 (33.7 pg·m − 3 ). These results indicate that air concentrations of PeCB in the city of Medellin increased by approximately four times in two years of monitoring.
On the other hand, concentrations in the air of DDT, DDE, and DDD isomers were constant during the monitoring time (Fig. 1a,1b). The air concentration range of Σ 4 DDTs was 213.5-269.1 pg·m − 3 ; with p,p'-DDE being the isomer with the highest levels in this family (109.6-145.0 pg·m − 3 , Table S3). In this respect, it is important to mention that DDT was banned in Colombia many years ago. Similarly, other . Trends for the other basic POPs were not identi ed due to their low concentrations in air, and in several cases were not detected at all (cis,trans-heptachlore epoxide, βendosulfan, endosulfan sulfate, β, δ-HCH-isomers, oxychlordane, cis-nonachlor, chlordecone, endrin, and mirex). All OCPs and indicators PCBs results are summarized in Table S3.
These results are in the same order of magnitude as those of other studies in Colombia and Latin America, which also suggests the persistent and global transport of basic POPs as DDTs, endosulfan, and heptachlor (Rauert et al., 2018a).

Dioxin like-polychlorinated biphenyls (dl-PCBs), Polychlorinated dibenzo-p-dioxins (PCDDs), and polychlorinated dibenzofurans (PCDFs)
dl-PCBs concentrations in air observed during the two years of monitoring are presented in Fig. 2a and Table S4. Finally, the PCDD/Fs ratios were calculated and are shown in Table S4. The PCDD/Fs ratios obtained were lower than 1.00 in the two years of monitoring, being 0.89 and 0.97 in 2017 and 2018, respectively. These results indicate a high contribution of PCDF congener. According to (Buekens et al., 2000), this pattern is possibly due to steel and iron furnaces. On the other hand, similar ratio values were found in areas of Manizales in uenced by industries such metallurgy, food, and plastic by (Cortés et al., 2014).
The ratio values found in this study suggest that PCDD/Fs concentrations in air are more associated with local sources in Medellin and could be associated with the metallurgy industries and steel and iron furnaces in the city.
This peak indicates punctual releases of these pollutants in this monitoring time, given that in the subsequent quartiles the air concentrations decrease signi catively. The principal PBDE congeners  Table S5.
PBDEs are commonly added to electronic products (shells, cables, and printed circuit boards), building materials, furniture, foam, textiles, and automobile parts, due to their ame are in the same order in magnitude; for example, 11.7 pg·m − 3 of Σ 9 PBDEs in Bogota was reported by (Saini et al., 2020). On the other hand, HBCDDs air concentrations in this study are higher than previous value reported by (Saini et al., 2020) in the which HBCDD isomers were not detected in Bogota. This suggests that the air concentrations of these compounds in Medellin are due to local sources.
Finally, PBB 153 was monitored in 2017 and 2018 but was not detected, the levels in the air being < 0.2 pg·m − 3 in all cases.

Per-and poly uoroalkyl substances (PFASs)
The PUF 2018-Q1 corresponding to PFASs was not able to quantify because the PUF was so deteriorated, thus, the data relating to the rst quarter of 2018 is not reported for these contaminants. On the other  On the other hand, the air concentrations of PFHxS and FOSA compounds did not change signi cantly in the two years of monitoring.
PFOS were in the same order of magnitude previously reported in the GRULAC region and country (Rauert et al., 2018b). The FOSA and FOSE air concentrations in this study were lower than those reported previously in the GRULAC region and the country (Rauert et al., 2018b). There are many sources of PFASs, for this reason, and due to the nature of this study, it is complex to associate the levels found to sources of release.

Conclusions
Air concentrations of a variety of POPs (chlorinated, brominated, and uorinated) were monitored in the atmosphere of Medellin, Colombia for two consecutive years. The results showed two punctual contamination peaks. The rst of these was associated with chlorinated compounds due to high releases of PeCB in the quartile 2017-Q3; emissions of PeCB are probably associated with unintentional formation in combustion processes in Medellin. The second contamination peak was associated with brominated compounds (PBDEs and HBCDD-isomers) and was observed in the quartile 2018-Q1. This contamination peak may be associated with recycling, incineration, and deposit at land ll of these compounds. On the other hand, the results suggest that PCDD/Fs concentrations in air are more associated with local sources in Medellin, such as the metallurgy industries and steel and iron furnaces in the city. Additionally, a slight tendency was observed for the air concentrations of PBDEs, HBCDDs, PCDDs, and PCDFs to increase in the year 2018 with respect to 2017. Furthermore, constant concentrations in air of pollutants banned in Colombia many years ago, such as DDT isomers and dieldrin, were observed, in contrast to a slight decrease in some pollutants such as aldrin, and α-endosulfan. Finally, the results provided a general vision of the air concentrations of chlorinated, brominated, and uorinated pollutants in Medellin, Colombia, and can be a starting point to nd the local sources of these pollutants and monitor their release in the city.

Declarations
Ethics approval and consent to participate Not applicable Consent for publication Not applicable Availability of data and materials All data generated and analyzed during this study are included in this published article and in the supplementary information les.

Competing interests
The authors declare that they have no competing interests

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. graphicalabstract.pptx Supportinginformation.docx