Article pubs.acs.org/est
Rapidly Equilibrating Micrometer Film Sampler for Priority Pollutants in Air Susan Genualdi and Tom Harner* Air Quality Processes Research Section, Environment Canada, 4905 Dufferin Street, Downsview, Ontario, Canada S Supporting Information *
ABSTRACT: Modified polymer-coated glass samplers (POGs), termed EVA samplers, consist of micrometer-thin layers of ethylene vinyl acetate (EVA) coated onto a glass fiber filter or aluminum foil substrate. These samplers were designed to equilibrate rapidly with priority pollutants in air, making them ideal for short-term spatial studies in ambient or indoor air. The EVA sampler was calibrated by measuring the uptake of polychlorinated biphenyls (PCBs) over 8 weeks in an indoor environment, and four different film thicknesses were monitored that ranged from 0.1 to 30 μm. The results were used to calculate the average mass transfer coefficient (50.5 m/ day) and generate contour maps that provide guidance in choosing an appropriate EVA sampler for a particular study based on film thickness, deployment time, and the log KOA of the anlayte. A range of air pollutant classes was also added to the EVA sampler prior to deployment to assess depuration rates. These included polychlorinated biphenyls (PCBs), current-use pesticides (CUPs), perfluorinated compounds (PFCs), and polybrominated diphenyl ethers (PBDEs). On the basis of the depuration profiles, the EVA sampler was a suitable equilibrium sampler for several CUPs and PCBs; however, for the high molecular weight PCBs and PBDEs, the EVA sampler operates as a linear uptake sampler. Samplers were also evaluated for their use as a rapid screening tool for assessing concentrations of siloxanes in indoor air. The EVA sampler was used to estimate air concentrations for D4 and D5 in laboratory air to be 118 and 89 ng/m3, respectively. Analyses were performed directly using thermal desorption gas chromatography/mass spectrometry (TDS-GC-MS). EVA samplers show promise due to their relatively low cost and ease of deployment and applicability to a wide range of priority chemicals. The ability to alter the film thickness, and hence the sorption capacity and performance of the EVA sampler, allows for a versatile sampler that can be used under varying sampling conditions and deployment times.
■
the partition coefficient between EVA and air.9 In order to calibrate kinetic samplers, compound-specific uptake rates need to be determined, typically through calibration experiments. Also, the sampler needs to be tested for varying environmental conditions that may be a factor at the sampling location such as temperature and wind speed.9 For equilibrium samplers, air concentrations are calculated by knowing the estimated partition coefficient of the target analytes between the sampler and air (i.e., the sorptive capacity of the sampler) and its temperature dependency. Examples of kinetic samplers include PUFs, SIPs, XAD, and SPMDs, whereas equilibrium samplers, which may also operate as kinetic samplers depending on the film thickness and analyte properties, include EVA, PE, SPME fibers, and POGs. The original polymer-coated glass (POG) sampler consisted of a 0.5 μm film of ethylene vinyl acetate coated onto a glass jar (68 mm o.d., 77 mm tall).5 This sampler has been calibrated and tested in both indoor and outdoor environments and has
INTRODUCTION Passive air samplers were designed to provide a simple, efficient, cost-effective alternative to active sampling for monitoring persistent organic pollutants in air. Many advances have been made in passive air samplers to address the need for a sampler that can measure air concentrations simultaneously on large spatial scales without the need for electricity. These samplers are continually being developed and modified to account for new chemicals of concern and applications. Examples of passive air samplers that have been developed include polyurethane foam disks (PUFs),1 polyurethane foam disks impregnated with ground XAD resin (SIPs), 2 XAD resin,3 semipermeable membrane devices (SPMDs),4 polymer (ethylene vinyl acetate) coated glass (POGs),5 ethylene vinyl acetate (EVA) and low-density polyethylene (LDPE) films,6 polyethylene devices,7 and solid-phase microextraction (SPME) fibers.8 Depending on the objective of the study, these samplers can be used as either kinetic or equilibrium samplers. Kinetic samplers can be used in monitoring studies to generate timeintegrated air concentrations over weeks or months. 9 Equilibrium samplers respond over shorter time periods such as hours or days, and the time to equilibrium is dependent on Published 2012 by the American Chemical Society
Received: Revised: Accepted: Published: 7661
April 11, 2012 June 7, 2012 June 17, 2012 June 17, 2012 dx.doi.org/10.1021/es301426s | Environ. Sci. Technol. 2012, 46, 7661−7668
Environmental Science & Technology
Article
both kinetic and equilibrium applications.5,10 Because the POG sampler has a high surface to volume ratio it reaches equilibrium rapidly in a matter of hours to days for most compound classes (5 days or less for compounds with log KOAs less than 9).5 Another advantage of this sampler is the ability to vary the film thickness. This allows for control of the time required to reach equilibrium, so that this can be optimized to meet sampling and analysis criteria. These criteria may include a specific sampling window (e.g., 2 days) and various instrument detection limits for particular analytes of interest. These samplers can be coated onto numerous substrates of varying size and shape and can be placed virtually anywhere. Compared to the SPME fiber, the POG sampler is inexpensive and has a larger surface area, resulting in a higher sampling rate, which is advantageous for rapidly measuring trace contaminants in air. This makes it an advantageous sampler for spatial studies involving numerous sites and for studies involving trace contaminants. The POG sampler can also be used to measure contaminants in water.11 A modified version of this sampler, the EVA sampler, is discussed here that uses glass fiber filters or thick aluminum foil as the substrate for ethylene vinyl acetate (EVA) instead of a glass jar. Because EVA is a copolymer between ethylene and vinyl acetate, it contains both nonpolar and polar functional groups, which are beneficial for absorption of polar and nonpolar compounds (e.g., CUPs and PCBs). In the current study, samplers of approximately 2.5 cm diameter were used. There is an increasing need for a sampler that can be applied to rapidly screen air for priority and emerging pollutant classes. The objectives of this research were to (1) calibrate the EVA sampler for its use as both a kinetic and an equilibrium sampler, (2) identify suitable analytes that can be measured using the EVA sampler, and (3) assess samplers for potential as a screening tool for priority air pollutants.
Coating of the EVA Sampler. A gastight syringe (Hamilton) was used to puncture a tiny hole in the EVA substrate. A syringe was used instead of tweezers to minimize handling of the EVA film. The syringe was further used to dip the substrates (aluminum foil and glass fiber filter) into the constantly stirred coating solution for about 5 s. Once removed from the coating solution, the sampler was spun around the syringe by gently pushing down on the edge of the sampler. This procedure was performed to remove excess solution and create an even coating. The sampler was dried quickly under a nitrogen stream (for about 30 s) and then placed on an aluminum foil-coated slide warmer (Fisher Sceintific) for about 1 min to fully dry. The samplers were then wrapped in bakedout aluminum foil and placed in the freezer until further use. Extra care was taken (e.g., prepared in a fume hood, fast processing time) to minimize the exposure time of EVA samplers to ambient air. Once everything was set up, making samplers was quite simple and fast, and within 1 h approximately 50 EVA samplers were made. EVA Film Thickness Calculations. In order to calculate the film thickness of the sampler created from each solution, the weights of 7 samplers were taken randomly throughout the coating process and the average weight of EVA added to each substrate was measured. Using this value, the density of EVA (g/cm3) and the surface area of the EVA sampler (10.12 cm2) the film thicknesses relating to the above solutions were calculated to be 0.1, 1.2, and 5 μm for the aluminum foil substrate EVA samplers and 2, 9, and 28 μm for the glass fiber filter substrate EVA samplers (Ex: weight (g)/density (g/cm3)/ surface area (cm3)). Even though the same EVA solutions were used for both aluminum foil substrate and glass fiber filter EVA samplers, the film thickness was consistently higher for the glass fiber filter (GFF) EVA samplers. The GFF is a relatively rough surface which is a porous weave of glass fibers. This results in greater retention of EVA during the coating process, whereas aluminum is a much smoother and impervious surface which retains less of the EVA. Substrate Comparisons. The glass fiber filters coated easily for the 2 and 9 μm film thicknesses. The 28 μm sampler proved more difficult in handling due to the stickier solution and additional time required for drying. For the aluminum foil substrate EVA samplers the thinnest film (1 μm) sampler coated easily; however, as the film thickness increased, it became more difficult to get the EVA solution to bind to the substrate and the sampler had to be handled carefully to avoid peeling. This issue was most likely due to the smooth and slippery surface of the aluminum foil; however, under further optimization aluminum foil could prove to be a promising substrate. For this study and method preparation scheme, aluminum foil is recommend for use when very thin EVA films are desired. Uptake Study. Three hundred EVA samplers (50 of each of the 6 different film thicknesses including both aluminum foil and glass fiber filter substrates) were deployed over a 142 day period (∼5 months) in an older and relatively unused laboratory by hanging them, from ceiling tile supports, approximately 5 cm below the ceiling. The laboratory was used in previous passive sampler calibrations and known to have elevated and consistent air concentrations of PCBs making it an ideal ‘chamber’ for long-term uptake studies.1 The temperature in the room was monitored over the deployment period and averaged 26 ± 1 °C. A low-volume air sampler was also deployed, and one PUF/XAD/PUF air sample was
■
METHODS Preparation of the EVA Substrate. The substrate for EVA samplers was created by taking a 2.5 cm circular punch of either glass fiber filter (VWR glass fiber filters grade 691) or a thick aluminum foil (VWR Disposable Aluminum Smooth-Wall Weighing Dishes). The punches were then baked out in a muffle furnace overnight at 400 °C prior to coating. Preparation of the EVA Coating Solutions. The ethylene vinyl acetate (EVA) coating solutions were created by adding a predetermined amount of EVA pellets (Elvax 40W, Dupont, Canada) to a glass jar along with a solvent mixture of 50:50 ethyl acetate:hexane. The jar was covered with aluminum foil, and the pellets were dissolved using a stir plate and a magnetic stir bar for about 2 h. Once the EVA pellets were dissolved, depuration standards were spiked into the solution and left to stir for another 30 min. The compounds added included 13C 10:2 FTOH (2-perfluorodecyl-(13C2)-ethanol), d3-N-MeFOSA (methyl-d3-perfluorooctane sulfonamide), d7-NMeFOSE (methyl-d7-perfluorooctane sulfonamidoethanol), d5N-EtFOSA (ethyl-d5-perfluorooctane sulfonamide, (Wellington Laboratories, Guelph, ON), PCB (polychlorinated biphenyls) 107, PCB 198, d 4 -Endosulfan I, 13 C trifluralin, BDE (polybrominated diphenyl ether) 25, BDE 75, BDE 116, BDE 181, and BDE 205 (Cambridge Isotopes, Andover, MA). Standards were spiked to generate the following EVA solution concentrations: 1, 4, and 8 g/100 mL or approximately 5 ng/ EVA sampler. 7662
dx.doi.org/10.1021/es301426s | Environ. Sci. Technol. 2012, 46, 7661−7668
Environmental Science & Technology
Article
collected every 2 weeks for the first 3 months of the study to get a total of 6 low-volume air samples. The sampler had a flow rate of 2020 cm3/min, and an average of 4070 m3 of air was collected over each 2 week period. Over the 5 month time period, EVA samplers were removed initially every few hours. The collection interval was increased in time over the study period until a sampler was being collected once per week. Once the EVA sampler was removed, it was placed in a small amber vial and placed in a freezer until further use. Three field blanks were collected for each film thickness. A field blank was taken by deploying an EVA sampler for a few minutes and then removing it and treating it like a sample. Analytical Method. A solvent-cleaned hole puncher was used to take a punch out of each EVA sampler for analysis. The hole punch was approximately 7 mm wide (1/4 in.) and represented approximately 8% of the total surface area of the EVA sampler. Therefore, approximately 10 punches can be taken from one EVA sampler to perform several different analyses if needed. However, if trace analysis is desired, several punches can be used at once and the entire EVA sampler can be thermally desorbed to obtain maximum sensitivity. The hole punch was placed inside a thermal desorption tube (Gerstel, Linthicum, MD), and an internal standard solution was spiked onto the filter punch. For internal standards, 13C endosulfan II was used along with unlabeled polybrominated diphenyl ether (BDE) 166, mirex, and pentadecafluoro-1-octanol (PDFO). These are all compounds not typically present in air. The TDS tube was then placed in a Gerstel TDS-A2 tray that was connected to a Gerstel thermal desorption system (TDS 3) and a cooled injection source (CIS 4). In the case of the more volatile PFCs, a modified tube was used for analysis. A glass wool plug was placed in a TDS tube followed by ∼0.5 g of Tenax TA, the EVA sampler was then inserted and spiked with internal standards, and then another 0.5 g of Tenax TA was added followed by another glass wool plug. This helped minimize the loss of anlaytes and minimize variability that occurred with these analytes and the thermal desorption system. The system was integrated with an Agilent 7890 GC and 5975 MS (Agilent Technologies). The instrumental parameters for CIS, TDS, and GC/MS for each type of analysis can be found in the Supporting Information (Table SI.1). QA/QC. EVA samplers that were deployed as field blanks were used to calculate concentrations at time zero for the native and labeled compounds. Four punches were taken from the same POG sampler to test the reproducibility of loading depuration compounds. The RSD of the average concentration measured in the replicate punches ranged from 1.2% to 14%. The percent difference was calculated using the theoretical amount initially spiked onto the sampler and the actual amount measured on the instrument. These values ranged from 4% to 22%, depending on the analyte.
by the amount of chemical in the air at equilibrium. If the partition coefficient is known, the concentration of a chemical in the air can be easily calculated. Octanol−air (KOA) partition coefficients have been previously measured for many POPs.12−18 Several studies have already investigated the linear relationship between the KEVA‑AIR partition coefficient and the octanol−air partition coefficient (KOA).5,19,20 In this study, KEVA‑AIR partition coefficients were calculated from the amounts of PCBs measured in the 2 μm film thickness sampler that had reached equilibrium divided by the concentrations of PCBs measured in air. In Figure 1, KEVA‑AIR derived in this study for
Figure 1. Regressions between the log KOA and the log KEVA‑AIR partition coefficients from previous studies and from this study.
PCBs at 26 °C shows good agreement compared with previous results for organochlorine pesticides (OCPs) and PCBs at 25 °C by Wu et al.,20 PCBs at 25 °C by Harner et al.,5 and OCPs and PCBs (20 °C) by Wilcockson and Gobas.19 Using this relationship, the log KOA of other similar chemical classes can be used to estimate the KEVA‑AIR partition coefficient. This can be further used to estimate air concentrations for chemicals that have come to equilibrium with the EVA sampler using eq 1 CA =
C EVA KEVA − AIR
(1)
If the analyte of interest does not come to equilibrium within the time frame required, the concentration in air can still be calculated using eq 2, which describes the full uptake profile C EVA = KEVA − AIR × CAIR ⎛ ⎛A ⎞⎞ kA × ⎜⎜1 − exp −⎜ EVA × ⎟t ⎟⎟ KEVA − AIR ⎠ ⎠ ⎝ VEVA ⎝
■
(2)
where CEVA is the concentration measured in the EVA sampler, KEVA‑AIR is the partition coefficient, CAIR is the concentration in air, AEVA is the area of the passive air sampler, VEVA is the volume of the passive sampler medium, kA is the air-side mass transfer coefficient, which needs to be calculated from the uptake study, and t is elapsed time.5 In order to calculate kA, the slope of the linear portion of the uptake curves is needed, along with the surface area of the EVA sampler and the concentration measured in air. Average air concentrations obtained through low-volume air samplers are reported in the Supporting Information (Table SI.2). During
RESULTS AND DISCUSSION Calibration of EVA Sampler. The ability of the EVA sampler to be either a kinetic or an equilibrium sampler is determined by the film thickness. Uptake is initially linear, and over time the uptake curve becomes curvilinear and then reaches equilibrium. At equilibrium, the amount of chemical sequestered by the EVA sampler is directly related to the partition coefficient (KEVA‑AIR), which is the amount (pg/m3 where m3 represents the volume of the EVA sampler) of chemical sequestered by the passive sampling medium divided 7663
dx.doi.org/10.1021/es301426s | Environ. Sci. Technol. 2012, 46, 7661−7668
Environmental Science & Technology
Article
Figure 2. Uptake profiles of 37 individual PCBs (tri to hepta) over 120 days into the EVA samplers at four different film thicknesses (0.1, 2, 9, and 28 μm).
appear to increase at first before decreasing and reaching equilibrium. This is most likely due to analytical variability, which is more pronounced in the curvilinear region of the uptake curve and more prevalent for low molecular weight compounds when using thermal desorption. Only four of the six original film thicknesses representing the greatest range of film thicknesses are presented, namely, 0.1, 2, 9, and 28 μm. Results for the 1.2 and 5 μm foil EVA samplers are excluded. The results in Figure 2 demonstrate the tendency for equilibration to occur rapidly for the thinner EVA films and for compounds with lower log KOA values. Depending on the analytes of interest and the type of sampler desired (equilibrium versus linear), an optimized EVA sampler can be made for a particular study by adjusting the film thickness to a desired equilibration or sample deployment time. Contour plots were made to illustrate these relationships and are shown in Figure 3. The contour plots can be used to help determine which EVA sampler (and film thickness) is best for a particular set of analytes over a given sampling interval. These plots were created using the time to 25% equilibrium (linear uptake sampler) and time to 95% equilibrium (equilibrium sampler) for PCBs. The rate constant for uptake into the EVA sampler (kU) was calculated using equation where δ is the film thickness in μm (eq 4).
the uptake study, 37 PCB congeners were detectable both in the lab air and in the EVA samplers. When possible, the slope was calculated from the linear uptake region and eq 3 was used to calculate kA kA =
slope AEVA CAIR
(3)
The average kA value over the 2, 9, and 28 μm film thicknesses was 50.5 m/day. Detailed results for each congener and film thickness are reported in Table SI.2, Supporting Information. This average kA value is lower but within a factor of 2 of values determined previously for PUF disk, SPMDs, and soil during a calibration study conducted in the same laboratory.1 We speculate that the lower uptake rate in this study may be related to the position of the EVA samplers near the ceiling where air flow is reduced, leading to slower air-side mass transfer.21 In the study conducted by Shoeib and Harner1 the passive sampling media were deployed in the middle of the room where they would have been subjected to higher rates of bulk air movement. Unfortunately, bulk air velocities in different parts of the laboratory were not measured during these two studies. Uptake of PCBs. PCB uptake curves for the EVA samplers can be found in Figure 2. Individual PCB congeners are represented by the dotted lines, and each dot along the line corresponds to an EVA sampler removal time. Some of the congeners in the uptake curves, particularly the tri-PCBs, 7664
dx.doi.org/10.1021/es301426s | Environ. Sci. Technol. 2012, 46, 7661−7668
Environmental Science & Technology
Article
Figure 3. Contour maps for EVA samplers at four different film thicknesses (0.1, 2, 9, and 28 μm) that can be used to identify which sampler is appropriate for a specific study based on the log KOA values of the compounds of interest and the deployment time.
kU =
kA 1 × δ KEVA − AIR
proportional to deployment time for a linear-phase sampler and to film thickness for an equilibrium sampler. Further details have been provided in Table SI.3, Supporting Information, related to the amount of individual PCB congeners collected by each of the four EVA samplers over time. Uptake of PFCs (Perfluorinated Compounds). Air concentration results for PFCs in the laboratory derived from the low-volume samplers are presented in Table SI.4, Supporting Information. These compounds include 6:2 FTOH, 8:2 FTOH, 10:2 FTOH, EtFOSEA, MeFOSEA, and MeFOSE. However, no uptake was detected for the thinner film (0.1, 1.2, and 5 μm) aluminum foil EVA samplers over the entire duration of the study, while the thicker film glass fiber filter EVA samplers showed detectable uptake into the 2, 9, and 28 μm EVA samplers after 3 days. Use of EVA samplers for measuring PFCs in air would therefore be limited to the thicker film EVA samplers that are able to retain greater amounts of PFCs. These detection issues for PFCs are attributed to the relatively low sorption capacity of EVA (i.e., low KEVA‑AIR value) for this class of compounds.
(4)
The time to 25% equilibrium can be further calculated by 0.29/ kU, and the time to 95% equilibrium was calculated using 3/kU. The contour plots show, as expected, that the 0.1 μm film thickness EVA sampler is fastest to reach equilibrium. With this sampler, compounds with log KOA values less than 8 equilibrate almost immediately. With increasing film thickness, the equilibrium region of the contour plot is reduced and limited to the low KOA region. For the higher KOA compounds, the thicker EVA film samplers can be operated as linear uptake samplers for short deployment times. Another operational or design consideration when choosing a particular film thickness and the deployment time is the sensitivity of the analysis method for the target analyte, i.e., how much chemical needs to be collected to achieve detection. In other words, it may not be advantageous to use a very thin sampler as this will limit the amount of chemical collected and the chances of detection. The amount of chemical collected is 7665
dx.doi.org/10.1021/es301426s | Environ. Sci. Technol. 2012, 46, 7661−7668
Environmental Science & Technology
Article
Figure 4. Depuration profiles of a select compound from 4 different compound classes (PFCs, pesticides, PCBs, and BDEs) at varying film thicknesses (0.1, 2, 9, and 28 μm).
Figure 5. EVA samplers as a screening tool. (A) D4 and D5 siloxanes because of their low log KOA values will enter the equilibrium region of the 9 μm sampler within 1 day. (B) Uptake curves for D4 and D5 show rapid equilibration with the 9 μm sampler in less than 10 days.
7.8), the EVA samplers experienced rapid depuration for all film thicknesses, indicating limited capacity of EVA for these compounds. For PCB 198 and BDE 205, which both have log KOAs greater than 11, the compounds remain in the EVA sampler and do not depurate, with the exception of the thinnest film (0.1 μm), which shows slight depuration over the 40 day time period. Complete results for the other depuration compounds are given in Figure SI.1, Supporting Information. The current-use pesticide, d4-endosulfan I, undergoes depuration in all four film thicknesses with the time to depletion increasing with film thickness. Since pesticides were not present in the laboratory air, the ability of pesticides to uptake into the EVA samplers was not observed. However, based on the depuration profiles of 13C trifluralin and d4endosulfan 1, EVA appears to be a suitable medium for this class of compounds and we expect that the nonpolar and hydrophobic current-use pesticides should behave similarly to PCBs and adhere to the log KEVA‑AIR − log KOA relationship
Depuration of Labeled Analytes. Isotopically labeled environmental contaminants and some unlabeled compounds that are not present in air were added to the EVA sampler spiking solution during preparation. Depuration compounds are used to provide additional information regarding the performance of passive samplers for a range of analyte classes and to correct for variability that may be due to meteorological conditions (e.g., wind speed, temperature). In this study, which took place in a controlled indoor environment, the depuration compounds were monitored to examine their loss rates from the EVA samplers. Depuration compounds from four different classes were measured: perfluorinated compounds (PFCs), pesticides, polybrominated diphenyl ethers (BDEs), and polychlorinated biphenyls (PCBs). Figure 4 shows examples of the depuration rates for one compound from each compound class for 4 different film thicknesses. A minimum of two points were used to create best-fit depuration lines for each labeled compound. For the labeled MeFOSA (log KOA ≈ 7666
dx.doi.org/10.1021/es301426s | Environ. Sci. Technol. 2012, 46, 7661−7668
Environmental Science & Technology
Article
samplers in multimedia environments (e.g., air−water-sediment). This would allow for estimation of chemical fugacities (relative fugacities) in these compartments simultaneously in order to gain insight into pollutant transport gradients and fluxes.
shown in Figure 1. This would need to be confirmed experimentally. The PCB 198 and BDE 205 depuration compounds, which both have high log KOA values (greater than 11), are highly retained by the EVA and show negligible loss over time (with the exception of PCB 198 in the 0.1 μm thin film). This confirms that for most compounds with high log KOA values the EVA sampler will be operating as a linear uptake sampler. EVA Samplers as a Screening Tool. It is desirable to have simple and rapid techniques to measure emerging pollutants in both indoor and outdoor air. These are tools that facilitate collection of data that can inform risk assessment, human exposure assessment, and source−receptor relationships. An example of a recent emerging environmental pollutant for which this technique may be used is the volatile methyl siloxanes (VMSs), in particular D4 (octamethylcyclotetrasiloxane) and D5 (decamethylcyclopentasiloxane), where 90% of their environmental loadings are released to the atmosphere.22 These compounds are present in personal care products such as lotions and deodorants and also in industrial applications. They have been measured in indoor dust23 and outdoor air (up to 280 ng/m3),22 with higher concentrations typically found in indoor environments. From Figure 3, any of the EVA samplers could be used to measure D4 and D5 siloxanes in the equilibrium region based on their log KOA values of 5.69, and 5.67.24,25 For the purposes of this study, the 9 μm sampler was used. The position of D4 and D5 on the contour plot for the 9 μm EVA film is shown in Figure 5A. Several 9 μm EVA samplers were analyzed from the study to generate the uptake curves shown in Figure 5B and demonstrate that equilibrium was established relatively quickly for both D4 and D5. These equilibrium concentrations in the EVA (CEVA) can be converted to air concentrations using the equilibrium expression in eq 1. The log KEVA‑AIR values for D4 and D5 can be obtained from the relationship from Figure 1 and the known log KOA values for D4 and D5. On the basis of this approach, the derived air concentrations for D4 and D5 in the laboratory were 118 and 89 ng/m3, respectively. These values are in the expected range based on measurements of D4 and D5 in indoor air (Mahiba Shoeib, personal communication). Blanks also need to be carefully investigated as background siloxanes are present in the column and septa (in this case, there was no septa because of the TDS/CIS inlet). A blank tube run prior to analysis showed that the siloxane response areas were 17% and 2.5% of the field blank EVA sampler response areas for D4 and D5. As demonstrated by this example, when equilibrium is achieved, calculation of air concentrations only requires knowledge of the KEVA‑AIR value, which can be estimated from log KOA. For future applications of the EVA sampler and to improve accuracy it may be desirable to measure and confirm KEVA‑AIR values for the target VMSs. Implications. EVA samplers are a promising tool for rapidly measuring pollutants in indoor and outdoor air. They can also be tailored and optimized to specific study objectives (e.g., sampling durations) and target analytes. The main advantage of these samplers is they can be rapidly analyzed after deployment and no sample preparation or extraction steps are needed. Of the two EVA-coating substrates investigated, aluminum foil had the advantage that it could provide very thin film samplers (
