Rain Amplification of Persistent Organic Pollutants

Scavenging of gas- and aerosol-phase organic pollutants by rain is an efficient wet deposition mechanism of organic pollutants. However, whereas snow has been identified as a key amplification mechanism of fugacities in cold environments, rain has received less attention in terms of amplification of organic pollutants. In this work, we provide new measurements of concentrations of perfluoroalkyl substances (PFAS), organophosphate esters (OPEs), and polycyclic aromatic hydrocarbons (PAHs) in rain from Antarctica, showing high scavenging ratios. Furthermore, a meta-analysis of previously published concentrations in air and rain was performed, with 46 works covering different climatic regions and a wide range of chemical classes, including PFAS, OPEs, PAHs, polychlorinated biphenyls and organochlorine compounds, polybromodiphenyl ethers, and dioxins. The rain–aerosol (KRP) and rain–gas (KRG) partition constants averaged 105.5 and 104.1, respectively, but showed large variability. The high field-derived values of KRG are consistent with adsorption onto the raindrops as a scavenging mechanism, in addition to gas–water absorption. The amplification of fugacities by rain deposition was up to 3 orders of magnitude for all chemical classes and was comparable to that due to snow. The amplification of concentrations and fugacities by rain underscores its relevance, explaining the occurrence of organic pollutants in environments across different climatic regions.

. PFAS, OPEs and PAHs sample recoveries of recovery standards (%) for rain and aerosols samples. Table S3. Limits of detection for PFAS, OPEs for rain and aerosols samples from Deception and Livingston s. The limits of detection (LODs) were defined as the mean concentration of field blanks plus three times the standard deviation of the blank response. For the analytes not detected in blanks, LOD were derived from the lowest standard in calibration curve. Table S4. Information of the data used in the meta-analysis of rain-air partition constants Table S5. KRP mean for each compound and for each data set in the meta-analysis. The compound order is the same as in Figure 1. Table S6. KRG mean for each compound and for each data set in the meta-analysis. The compound order is the same as in Figure 2. Table S7. KRA mean for each compound and for each data set in the meta-analysis. The compound order is the same as in Figure 3. Table S8. Dimension-less Henry's Law constant values and their sources used for the meta-analysis of KRGH'. Table S9. PFAS concentrations in rain (pg L -1 ) and aerosols (pg m -3 ) samples from Livingston and Deception Island. *= Data from Casas et al. 2020.i Date = start of the rain event or start of aerosol sampling. f Date = end of the rain event or end of aerosol sampling. Table S10. OPE concentrations in rain (pg L -1 ) and aerosols (pg m -3 ) samples from Livingston Island. i Date = start of the rain event or start of aerosol sampling. f Date = end of the rain event or end of aerosol sampling. Table S11. PAH concentrations in rain (pg L -1 ) and aerosol (pg m -3 ) samples from Livingston Island. i Date = start of the rain event or start of aerosol sampling. f Date = end of the rain event or end of aerosol sampling.

S3
Internal standards were spiked before all instrumental analysis for quantification and to evaluate instrument performance (Table S1).
The chromatographic separation for OPEs was carried out using an Agilent HP-5MS column (30 m, 0.25 mm internal diameter, 0.25 µm film thickness) by an Agilent 7890 GC. Methane was used as ionization gas and helium was used as a carrier gas at a constant flow mode at 20ml min -1 . Two µL of sample were injected in split less mode and the injection port temperature was 280 °C. The column temperature ramp for an effective separation of analytes were as follows; 90 °C for 1 min, increased at 15 °C min -1 to 200°C and held for 6 min, then at 5 °C min -1 to 250 °C, and held for 6 min and then at 10°C min -1 to 315°C, and held for 10 min. The detection was carried out with an Agilent 7000B triple quadrupole mass spectrometer using electro impact ionization (EI) mode in positive conditions. The EI source, transfer line and quadrupole temperature were 230°C, 280 °C and 300 °C, respectively. Acquisition was performed in MRM.
The chromatographic separation for PAHs was carried out using an Agilent DB-5MS column (30 m, 0.25 mm internal diameter, 0.25 µm film thickness) by Agilent 7890 GC.
Two µL of sample were injected in split less mode. The initial GC oven temperature was set at 90°C. It was risen up to 175°C at a rate of 6 °C min -1 and held for 4 min. Then the temperature increasing rate slowed to 3 °C min -1 until 235 °C. After, the heating rate was switched to 8 °C min -1 and held for 8 min, until 300 °C. At last, with the same rate, the temperature reached 315 °C during 4 min. The detection was carried out with an Agilent 5975C mass spectrometer using EI mode in positive conditions. Acquisition was performed in selected ion monitoring (SIM).
Annex S2. Quality Assurance/Quality Control All recipients, tubes and connections used from the sampling to the chemical analysis of OPEs and PAHs were made of stainless steel, glass or PTFE. Nevertheless, for PFAS analysis all recipients were made of stainless steel or PP. These were pre-cleaned with methanol and acetone prior use in order to avoid contamination. All filters were precombusted at 450°C over 4h.
Procedural blanks, consisting of QFFs and SPE cartridges, were processed analogously to samples. In addition, field blanks consisted of QFFs and cartridges that were transported to sampling sites, shipped back to the laboratory with samples, processed in the same manner as samples albeit without the pass of rain water or air. Recovery of surrogate standards spiked before the extraction for the different types of samples and for each compound family are presented in Table S2. The limits of detection (LODs) were defined as the mean concentration of field blanks plus three times the standard deviation S7 of the blank value. For the analytes not detected in blanks, LOD were derived from the lowest standard in calibration curve. Limits of detection are presented in Table S3.
Annex S3. Uncertainty error propagation estimation for KRG/KSA. Where is the standard error. Figure S1. Sampling location for the rain and aerosol samples analysed in this study.    Figure S5. Meta-analysis of rain-air particulate partition constants (KRP) for various families of organic pollutants differentiating the type of aerosol (Continental, Coastal, Open ocean, Urban). The results shown are the mean and the standard deviation of log KRP. Figure S6. Meta-analysis of rain-air, gas phase adsorbed (KRG, adsorbed) for various families of organic pollutants. KRG, adsorbed as given by Equation [6], = , + , , where KRG,dissolved is 1/H'. The results shown are the mean and the standard deviation of log KRG,adsorbed. S14 Figure S7. Pearson's correlations between log KRA versus log Kaw, log Koa, log Kow. S15 Figure S8. Pearson's correlations between log KRG versus log Kaw, log Koa, log Kow. Figure S9. Pearson's correlations between log KRP versus log Kaw, log Koa, log Kow.   Table S5. KRP mean for each compound and for each data set in the meta-analysis. The compound order is the same as in Figure 1.  Table S6. KRG mean for each compound and for each data set in the meta-analysis. The compound order is the same as in Figure 2.  Table S7. KRA mean for each compound and for each data set in the meta-analysis. The compound order is the same as in Figure 3.