Neutral Polyfluoroalkyl Substances in the Global Atmosphere † Environmental Impact

Concentrations of neutral per-and polyfluoroalkyl substances (nPFAS) in the atmosphere are of interest because nPFAS are highly mobile percursors for perfluoroalkyl acids. Two calibration studies in Ontario, Canada and Costa Rica established the feasibility of using XAD 2-resin based passive air samplers (XAD-PAS) to reliably determine long term average air concentrations of nPFAS under temperate and tropical climatic conditions. The temporal and spatial distribution of nPFAS was investigated by analyzing XAD-PAS deployed for one year at between 17 and 46 sites on six continents between 2006 and 2011 as part of the Global Atmospheric Passive Sampling (GAPS) study. Higher levels of fluorotelomer alcohols (FTOHs) compared to fluorinated sulfonamides (FOSAs), and fluorinated sulfonamidoethanols (FOSEs) were observed at all sites. Urban sites had the highest levels of nPFAS compared to rural and remote sites, which is also apparent in a positive correlation of nPFAS levels with the proximity of a sampling site to areas of high population density. Levels of FOSAs and FOSEs tended to decrease during the six years of measurements, whereas an initial decline in the concentrations of FTOHs from 2006 to 2008 did not continue in 2009 to 2011. A comparison of nPFAS levels measured in national XAD-PAS networks in Costa Rica and Botswana revealed that the GAPS sites in Tapanti and the Kalahari are representative of the more remote regions in those countries. XAD-PAS derived absolute nPFAS levels at GAPS sites are lower than those measured using another PAS, but are within the range of levels measured with active air samplers. Agreement of relative nPFAS composition is better between samplers, suggesting that the discrepancy is due to uncertain sampling rates. Peruoroalkyl acids have been receiving widespread attention due to their persistence, ability to bioaccumulate and ubiquitous presence in the global environment. Some volatile neutral per-and polyuoroalkyl substances (nPFAS) have been implicated in facilitating the transfer by atmospheric long range transport to remote regions, where they may degrade into the peruoroalkyl acids. This study for the rst time provides insight into interannual time trends of nPFAS in the atmosphere on a global scale and conrms their ubiquitous presence throughout the global atmosphere.


S2
Table S1 nPFAS analytes, acronyms, and target and qualifier ions for GC/MS detection.S4 Table S2 Concentrations of nPFAS sequestered on XAD-PAS.S5 Table S3 Statistical results for the sampling uptake curves S5 Table S4 Concentrations and MDLs of selected nPFASs in active air samples (AAS).S6 Fig. S1 Seasonal variability of nPFAS concentrations in the 3 AAS campaigns.S8 Table S5 Results from 1-way ANOVA test for nPFAS concentrations between the three active air-sampling sites.S8

Table S6
Name, geographical coordinates (latitude, longitude), site classification (AG: Agricultural, BA: Background, PO: Polar, RU: Rural, and UR: Urban), and length of sampling period for GAPS sites

Table S7
Concentrations and method detection limits (MDL) in ng•PAS -1 of selected neutral polyfluoroalkyl substances (nPFAS) at GAPS sites.S11

Table S8
Percentage of individual nPFAS detected per year with number of sites for that year.S16

Fig. S2
Relative abundance of nPFAS on a global scale.S16 Fig. S3 Temporal trends at the 13 sites that were consistently part of the GAPS program.S17 Table S10 Results from 1-way ANOVA test investigating the differences in nPFAS concentrations between site types.S17 Table S9 1-way randomized block design ANOVA result for the concentrations of individual nPFAS.S18

Fig. S4
Pearson correlation between the concentrations of the nPFASs in XAD-PAS and the pertingency index of the sampling site S20

Text
Regional differences in nPFAS Levels.S21 Fig. S5 Box and whiskers plots for nPFAS separated according to the four major world regions and proximity to emissions.S22

Fig. S6
Temporal trends across the four global regions.S23 Table S11 1-way ANOVA on nPFAS concentrations based on region.S24 Table S12 nPFAS correlations among all samples in the global environment and region S25

Sampling Site Descriptions
Twenty-seven, 24-hour high volume AASs were collected in Maun, Botswana every two weeks between May 2006 and May 2007. 1 During that one year period, duplicate XAD resin-based PASs were deployed at ten sites within or near to the Okavango Delta in Northwestern Botswana, with five more in the more populated East of the country. 1 Located in southern Africa, Botswana's climate is semiarid with consistently high temperatures throughout the year.Rainfall varies, being heaviest in the Northeast (~650mm) and sparse (~<250mm) in the Kalahari Desert to the Southwest.Detailed description of the sites is given in Figure 1 and in ref. 1 Eight 24-hour high volume AASs were taken in San Antonio de Belen in Costa Rica's heavily populated central valley between October 2005 and October 2006.At the same time, duplicate XAD resin-based PASs were deployed across the country at seven sites in remote or agricultural areas. 2 Concurrently, five duplicate sets of PAS collected air in Belen for variable lengths of time, ranging from four months to somewhat more than one year, in order to determine the uptake of nPFAS in XAD-2 based samplers in tropical zones. 3Costa Rica, a mountainous Central American country situated near the equator, has a tropical climate with consistently high temperatures throughout the year and a distinct rainy season.activity are influential factors for the presence of nPFAS in the atmosphere [12][13][14] sites categorized as 'agricultural', 'rural', or 'background' were reclassified as either 'remote' or 'urban', so that a simplified categorization only distinguished between urban, remote, and polar sites (Table 1, Figure 1).For geographic comparisons, sites were also grouped into four regions: 5 North America (9-18 samples), South (5-11), Europe (5-8), and Asia (2-12) (Figure 1).

Sample Preparation and Collection
Preparation and collection of the samples have been described in detail elsewhere. 15In the case of PAS, 10 cm (Botswana, Ontario) or 20 cm (GAPS, Costa Rica) long mesh cylinders (diameter: 2 cm) were filled with pre-cleaned XAD-2 resin (Supelpak 2, Supelco, Bellefonte, PA).Mesh cylinders were placed into stainless cylinder tubes closed with stoppers, sealed in plastic bags, and shipped to the sites.Field blanks were treated in the same manner as the samples, except that the blanks were not exposed to air.The high volume AAS used in the Costa Rica and Botswana campaigns consisted of 10 g XAD-2 sandwiched between two large (width wPUF 6 cm, length lPUF 3 cm) PUF plugs.The low volume AAS used in the Egbert campaign was comprised of 5 g XAD-2 between two small (wPUF 2 cm, lPUF 3 cm) PUF plugs.
Samples from Botswana, 1 Costa Rica, 3,2 Ontario 16 and 2006-2008 GAP samples 5 underwent Soxhlet extraction overnight with dichloromethane.These samples had not been spiked with isotope labeled nPFAS prior to extraction.Recoveries of laboratory spikes (n=6) were on average between 77-96%, 111-124% and 113-115% for FTOHs, FOSAs and FOSEs, respectively.As these recoveries were judged acceptable, samples were not recovery-corrected, aside from adjustment for any detector response differences during gas chromatography-mass spectrometry (GC-MS) determination.This correction was made by spiking 5-12 ng of isotope labeled nPFAS prior to analysis on the GC-MS.
2009-2011 samples within GAPS were extracted using pressurized liquid extraction (Dionex ASE ® 350) with 1:1 (v/v) acetone:hexane at 75 °C, adapted from Dionex's protocol. 17Primbs et al. 18  After extraction, all samples were volume reduced using either a rotary evaporator or Turbovap, further concentrated to 500 l under nitrogen, and then solvent-exchanged into iso-octane.Aside from passing the samples through ~1.0 g of sodium sulfate to remove moisture, no clean-up was done on the extracts.Additionally, prior to injection, 10 ng of N,N-Me2FOSA was added into the samples for volume correction.
Analysis of the samples was performed using GC-MS in selected ion monitoring mode using positive chemical ionization (ions in Table S1).Aliquots of 2 L were injected and separated on a 30 m DB-WAX column with a 0.25 mm inner diameter and 0.25 μm film thickness.Helium was the carrier gas at a flow of 1.2 mL•min -1 .Temperature program was as follows: 60-65 °C (held for 3 min), ramped at 2 °C•min -1 to 70 °C, then ramped at 8 °C•min -1 to 120 °C, and finally ramped at 10 °C•min -1 to 220 °C with a post run at 230 °C for 3 min.A 5-10 point calibration curve (0.48-96 pg•l -1 ) was used for quantification, using the isotope dilution method.

Regional Differences in nPFAS Levels
Figure S3 compares the levels of the seven analytes in four world regions.Given that nPFAS levels were correlated with proximity to people, sites are separated based on region and site type (remote vs. urban).Because of their limited number, the urban sites can hardly be considered representative for all urban areas in a region.Polar sites were not included in this analysis at all because of their small number.At the remote sites, the order of total nPFAS concentrations from lowest to highest generally is: South < North America < Europe < Asia.Lower nPFAS levels in the southern hemisphere might be expected, considering that use of nPFAS is much lower. 24On the other hand, Asian countries, namely China, still manufacture PFOS and its derivatives, including perfluorooctane sulfonyl fluoride. 25However, statistically, none of the differences between the four regions is significant (Table S11).
Pearson correlation statistics of individual compounds was conducted for the global environment and also for the four regions (Table S12).It was suggested previously that correlations between compounds imply similar sources and fates. 26,27In North America, correlations were highly significant (R>0.61,p<0.01) for all compounds, in Asia and Europe, approximately half of the compounds correlate with one another, whereas in the South correlations between the nPFASs were weakest.Aside from EtFOSA, all the compounds were correlated with one another on a global scale.As the emissions of all of the analytes are concentrated in urban areas, the correlations are likely just another indication that sites with high PI have elevated nPFAS levels.At a regional scale (Figure S6), both North America and Europe showed a remarkable nPFAS decrease between 2006 and 2007.After 2007, levels increase again, although not as high as those measured in 2006.Ahrens et al. 28 have observed an increase in FTOH amounts in the Canadian Archipelago between 2005 and 2007/2008.Interestingly, efforts since 2005 have been made to reduce fluorotelomer-based residuals in Canada 29 and the United States 30 .However, due to long product lifetimes, it will take some time for households to switch to newer products that contain fewer FTOH residuals. 31Levels of all FTOHs in the South have remained steady, whereas in Asia, FTOH levels have been declining, although not significantly.
As for the FOSAs and FOSEs, there are no observable declines, aside from North America where after 2006 levels have dropped significantly (p<0.05,Table S9).The phase-out of perfluorooctane sulfonates and related compounds 25

Figure S1 .
Figure S1.Seasonal variability of nPFAS concentrations during the 3 active air-sampling campaigns.Note the different scales on the y-axis.Blue shading indicates rainy seasons in Costa Rica and Botswana.

Figure S3 .
Figure S3.Temporal trends (2006-2011) of nPFASs in XAD-PAS at the 13 sites that were consistently part of the GAPS program.

Figure S4 .
Figure S4.Pearson Correlation between the concentrations of the nPFASs in XAD-PAS and the pertingency index of the sampling site, which expresses the proximity to people.In the case of the plot for sum of nPFAS, levels below the IDL were assigned values between 1/3-2/3 of IDL, which adds uncertainty to this regression.

Figure S5 .S23Figure S6 . 1 ]
Figure S5.Box and whiskers plots for nPFAS separated according to the four major world regions and proximity to emissions (i.e."remote" and "urban").Polar sites were not included due to their small number.

Table S13
Average nPFAS concentrations in duplicate XAD-PAS in national PAS campaigns S26 [4][5][6][7][8]The data will aid in evaluating the effectiveness of the restrictions imposed on POPs under the Stockholm Convention.Two types of PASs are used in GAPS; polyurethane foams (PUFs) or sorbent-impregnated PUFs (SIPs) are deployed at a seasonal resolution, whereas XAD-PASs yield annual averaged air concentrations.Specifically, such PASs were deployed at 34, 46, 33, 34, 22, and 17 sites during each of six sampling years, which roughly correspond to calendar years 2006 to 2011.Samples from the first year of the GAPS study, in 2005, were not included because extract clean-up with alumina columns compromised the analysis for the analytes of interest.Because previous sampling campaigns showed generally good agreement between replicates

Table S3 .
Statistical results for the sampling uptake curves Electronic Supplementary Material (ESI) for Environmental Science: Processes & Impacts This journal is © The Royal Society of Chemistry 2013
Electronic Supplementary Material (ESI) for Environmental Science: Processes & Impacts This journal is © The Royal Society of Chemistry 2013 Part 2: Egbert, Ontario, Canada

Table S5 .
Results from 1-way ANOVA test for nPFAS concentrations between the three active air-sampling sites.Comparisons between sites used Tukey-Kramer's Multiple Comparison Test.NS = not significant.] Maun, Botswana (HiVol, 24 hour each) San Antonio de Belen, Costa Rica (HiVol, 24 hour each) Egbert, Ontario, Canada (LoVol, 2 week each) Electronic Supplementary Material (ESI) for Environmental Science: Processes & Impacts This journal is © The Royal Society of Chemistry 2013

Table S2 .
Name, geographical coordinates (latitude, longitude), site classification (AG: Agricultural, BA: Background, PO: Polar, RU: Rural, and UR: Urban), and length of sampling period for GAPS sites that had XAD-PAS deployed between 2006 and 2010.Electronic Supplementary Material (ESI) for Environmental Science: Processes & Impacts This journal is © The Royal Society of Chemistry 2013

Table S3 .
5oncentrations and method detection limits (MDL) in ng•PAS -1 of selected neutral polyfluoroalkyl substances (nPFAS) at GAPS sites.Values were normalized to 365 days of sampling.ND = not detected (below instrument detection limit, IDL); BDL = below MDL.Questionable concentrations, which were not included in the figures and statistical analyses have * next to the numbers.Additional sampling site information can be found in TableS1in Shunthirasingham et al.5First sampling year (late 2004 to late 2005) not included due to loss of analytes.

Table S4 .
Percentage of individual nPFAS detected per year with number of sites for that year.Relative abundance of nPFAS on a global scale.

Table S5 .
Results from 1-way ANOVA test investigating the differences in concentrations between site types.

Table S6 .
1-way randomized block design ANOVA result for the concentrations of individual nPFAS between years 2006-2011.P-values <0.05 suggest nPFAS concentrations are significantly different between sampling years.Between two years, used post-test to compare statistically significance within the five years of sampling (post-test using Tukey-Kramer's Multiple Comparison Test.).NS = not significant.
25curred already in 2000-2002 in most parts of the world (aside from China25and Brazil 32 ).It thus comes as no surprise that the FOSA and FOSE concentrations are near detection limits and show no significant decline during the period of sampling.

Table S7 .
1-way ANOVA on nPFAS concentrations based on region.This analysis only considered sampling sites located in remote locations.6

Table S9 .
Average concentrations of nPFAS in duplicate XAD-PAS [ng.PAS-1] deployed across Costa Rica (Part 1) and Botswana (Part 2).Samples were normalized to 365 days of sampling.MDL = method detection limit; ND = not detected (below instrumental detection limit); BDL = below MDL.Additional sample site information for Botswana can be found in TableS1in Shunthirasingham et al. 1 .Electronic Supplementary Material (ESI) for Environmental Science: Processes & Impacts This journal is © The Royal Society of Chemistry 2013 Electronic Supplementary Material (ESI) for Environmental Science: Processes & Impacts This journal is © The Royal Society of Chemistry 2013 * Detected in one of the two samples