Influence of Temperature on the Emission of Di-(2-ethylhexyl

Dec 16, 2011 - and five different temperatures: 23 °C, 35 °C, 47 °C, 55 °C, and 61 °C. The experiments were terminated two weeks to three months ...
0 downloads 0 Views 1MB Size
Download PDF
Article pubs.acs.org/est

Influence of Temperature on the Emission of Di-(2-ethylhexyl)phthalate (DEHP) from PVC Flooring in the Emission Cell FLEC Per Axel Clausen,†,* Zhe Liu,‡ Vivi Kofoed-Sørensen,† John Little,‡ and Peder Wolkoff† †

National Research Centre for the Working Environment, Lersø Parkalle 105, DK-2100 Copenhagen Ø, Denmark Department of Civil and Environmental Engineering, Virginia Tech, Blacksburg, Virginia 24061-0246, United States

‡

S Supporting Information *

ABSTRACT: Emissions of di-(2-ethylhexyl) phthalate (DEHP) from one type of polyvinylchloride (PVC) flooring with approximately 13% (w/w) DEHP as plasticizer were measured in the Field and Laboratory Emission Cell (FLEC). The gas-phase concentrations of DEHP versus time were measured at air flow rate of 450 mL·min−1 and five different temperatures: 23 °C, 35 °C, 47 °C, 55 °C, and 61 °C. The experiments were terminated two weeks to three months after steady-state was reached and the interior surface of the FLECs was rinsed with methanol to determine the surface concentration of DEHP. The most important findings are (1) DEHP steady-state concentrations increased greatly with increasing temperature (0.9 ± 0.1 μg·m−3, 10 ± 1 μg·m−3, 38 ± 1 μg·m−3, 91 ± 4 μg·m−3, and 198 ± 5 μg·m−3, respectively), (2) adsorption to the chamber walls decreased greatly with increasing temperature (measured partition coefficient between FLEC air and interior surface are: 640 ± 146 m, 97 ± 20 m, 21 ± 5 m, 11 ± 2 m, and 2 ± 1 m, respectively), (3) gas-phase DEHP concentration in equilibrium with the vinyl flooring surface is close to the vapor pressure of pure DEHP, and (4) with an increase of temperature in a home from 23 to 35 °C, the amount of DEHP in the gas- and particle-phase combined is predicted to increase almost 10-fold. The amount in the gas-phase increases by a factor of 24 with a corresponding decrease in the amount on the airborne particles.

■

INTRODUCTION Since the 1930s, phthalates have been used as plasticizers to enhance the flexibility of rigid polyvinylchloride (PVC) products, with worldwide phthalate production in 2004 of about 6 million tons/year.1 Di(2-ethylhexyl) phthalate (DEHP) is most widely used and may be present at concentrations as high as 10−60% (w/w).1 Phthalates are not chemically bound to the polymer matrix, thus they are slowly emitted from the products to air or other media. As a result, phthalates are ubiquitous and among the most abundant semivolatile organic compounds (SVOCs) in indoor environments (e.g., refs 2 and 3). Several studies4−7 suggest that exposure to phthalates increases prevalence of asthma, rhinitis, or wheezing in children (although this has been questioned8,9), causes reproductive disorders in humans, and affects endogenous hormones.1 Emission of DEHP from PVC is controlled by mass transfer in the boundary layer at the surface10−12 and is not influenced by humidity of the air.11,13 Emission measurements in stainless steel chambers such as the Field and Laboratory Emission Cell (FLEC) is strongly influenced by the fact that ca. 50% of the emitted DEHP is sorbed to the internal steel surfaces of the test chamber at ambient temperature.10 The emission of DEHP from vinyl flooring in the FLEC can be reasonably predicted by a fundamental mass transfer model.14 However, this model requires a well-mixed chamber, which is not the case for the FLEC. Therefore, computational fluid dynamics (CFD) was © 2011 American Chemical Society

used to account for the nonuniform concentration distribution of SVOCs in the FLEC.12 Although needed for risk assessment and control strategies, the influence of temperature on emission and distribution of phthalates in indoor environments is still inadequately investigated. Kim et al.15 showed that a temperature increase from 30 to 60 °C resulted in a 2−3 fold increase in the steady state concentration of DEHP in water/ethanol (55:45 by volume) during migration from plasticized PVC. Fujii et al.16 investigated temperature dependence of the emissions of various phthalate esters in a passive flux sampler. They observed a 100-fold increase in the maximum emission rate with an increase in temperature from 20 to 80 °C. The diffusion sampler was placed within 0.5 or 2.0 mm from the source without exchange of air. Since the phthalate concentration at the diffusion sampler surface is approximately zero, this results in a very short and steep concentration gradient and thereby a very high emission rate. Ekelund et al.17 studied aging of PVC plasticized with DEHP by measuring weight loss over time at temperatures from 60 to 155 °C. They concluded that evaporation (diffusion in the boundary layer) controlled emissions below 100 °C. The evaporation of DEHP Received: Revised: Accepted: Published: 909

October 7, 2011 December 12, 2011 December 16, 2011 December 16, 2011 dx.doi.org/10.1021/es2035625 | Environ. Sci. Technol. 2012, 46, 909−915

Environmental Science & Technology

Article

Table 1. Measured Temperatures, Surface and Steady-State Concentrations of DEHP in the FLECs and Measured Ksteel as Function of Temperaturea T (°C) 23.0 ± 0.4 (22.1; 25.5) 23.0 ± 0.4 (21.9; 25.5) 35.3 ± 0.1 (35.0, 35.6) 47.4 ± 0.1 (46.7; 47.6) 55.3 ± 0.5 (53.9; 56.8) 61.1 ± 0.5 (59.4; 62.6)

Csurf (surface conc.)b (μg·m−2) 608 588 949 798 1043 456

± ± ± ± ± ±

181 181 181 181 181 181

Csteady (Steady-state conc.) (μg·m−3) 1.0 0.9 10 38 91 198

± ± ± ± ± ±

0.1 (0.9; 1.6) 0.1 (0.5; 1.3) 1 (6; 18) 1 (36; 40) 4 (83; 104) 5 (184; 210)

Csteady -time interval of measurements (days)

Ksteelc (m)

16−97 16−97 6−85 4−18 2−18 1−28

599 ± 187 682 ± 224 97 ± 20 21 ± 5 11 ± 2 2±1

a The measures are given as mean ±95% confidence limit (min; max). The two experiments at 23 °C were run in parallel. bConfidence limit based on pooled standard deviation estimated on duplicate analysis of extracts rinsed off the FLEC surfaces. cConfidence limit based on error propagation from Csurf and Csteady.

below 100 °C had an activation energy of 98 kJ/mol, which was equal to that for evaporation of pure liquid DEHP. The experiments were carried out in ventilated ovens with no control of the air exchange rate, which is known to be of importance for the emission of DEHP from PVC.12 The objectives of the present study were to determine the effect of temperature on emissions of DEHP from one type of vinyl flooring in the FLEC using both experimental and CFD modeling techniques and to investigate the relationship between emission characteristics and the vapor pressure of DEHP. Finally, a recently developed model18 which predicts DEHP concentration in a household was used to demonstrate the impact of temperature on DEHP concentrations in a more realistic residential environments.

The emission rate of DEHP from vinyl flooring is not influenced by the humidity,13 thus ∼0% RH was used to avoid humidification of the dry compressed air. Before the experiments the FLECs were cleaned by wiping with ethanol and water and finally dried in a vacuum oven at 100 °C for 2−3 h. The 23 °C experiment, in duplicate, and a simultaneous blank test (FLEC placed on a cleaned glass plate) were performed in a temperature-controlled laboratory. For all of the other temperatures, the FLECs, including samples and sample stands, were placed in incubators (Memmert INE 200, Germany, and Heraeus B 20, Fishers Scientific, Germany). The air concentration of DEHP, temperature, humidity, and flow rate in the outlets of all FLECs was measured simultaneously at regular intervals. The background concentration in the blank FLEC was monitored over a 247 day period with 60 duplicate measurements. Before the experiments the background concentration in all FLECs was measured. The temperature and humidity was measured with a Testo 650 instrument (Testo, Lenzkirch, Germany). The air flow rates were measured with a traceable electronic soap bubble flow meter (“The Gilibrator” (primary flow calibrator), Gilian Instrument Corporation, Caldwell, NJ, U.S.). The experiments were terminated two weeks to three months after steady-state was reached. The interior surfaces of all FLECs were rinsed with methanol to estimate the surface concentrations of DEHP, Csurf, as described below. Sampling and Analysis of DEHP in FLEC Air. During emission of DEHP in the FLECs the air is considered to contain negligible amounts of airborne particles and gas-phase DEHP was sampled in duplicate on Tenax TA with backup tubes to check for breakthrough as previously described12,13 and analyzed by TD-GC-MS. No breakthrough was observed in any of the samples. Before analysis, 5 μL solution of D4-DEHP in methanol was injected into the Tenax tubes as internal standard. The sampling time was usually 15−60 min corresponding to 3−10 L sampling volume. However, for the low concentrations of the 23 °C experiments sampling time was 1097−8570 min corresponding to 217−1831 L. The flow controlled sampling pumps (Gillian 5 from Gillian or FLEC Air Pump 1001 from CHEMATEC) were calibrated to a nominal flow rate of 200 mL·min−1 and flow checked at 23 °C before and after each sampling with “The Gilibrator”. The difference of the flows before and after sampling was less than 2%. During sampling, pumps were kept outside the incubators while the Tenax sampling tubes were inside at the elevated temperature. Sampling tubing between Tenax tubes and pumps was sufficiently long to cool the sampled air to 23 °C. Because of the contraction of the sampled air at the lower temperature the measured concentrations were corrected for the temperature

■

EXPERIMENTAL SECTION Chemicals. DEHP (99.5% (GC)) for calibration and identification was from Riedel de Haën. Tetra deuterium ring labeled DEHP (98%) (D4-DEHP) used as internal standard was from Cambridge Isotope Laboratories, Inc. Acetone (pro analysi, 99%) and methanol (Lichrosolv 99.6%) used for cleaning and as solvent in the standard solutions were from Merck. Test Piece. The test pieces (0.25 × 0.25 m) were vinyl flooring (according to the manufacturer homogeneous polyurethane reinforced PVC flooring) with a thickness of 2.0 mm which had been used in previous experiments.12,13 The content of DEHP was estimated as 13% ± 2% (w/w; mean ±95% confidence limit) by pressurized liquid extraction (Dionex ASE 200) of 14 samples, which were cut out with a scalpel to ca. 20mg pieces (ca. 1 mm × 2 mm × 10 mm) at random locations in one piece. The samples were extracted in 5-mL cells with methanol at 150 °C and 14.0 × 106 Pa (heating for 9 min, static extraction for 10 min, and flushed by 25% of volume). Aliquots (5 μL) of the extracts were injected together with 5 μL of a solution of D4-DEHP as internal standard into Tenax TA sorbent tubes and analyzed by thermal desorption combined with gas chromatography and mass spectrometry/flame ionization (TD-GC-MS/FID), as previously described.12,19 Validation of the TD-GC-MS method by on-column injection resulted in a mean value not significantly different, albeit with a larger variation (14% ± 3%). Other phthalate esters were detected only in trace amounts. The Emission Experiments. The concentrations of DEHP versus time were measured in FLECs (Field and Laboratory Emission Cell, CHEMATEC, Denmark)20,21 at an air flow rate of 450 mL·min−1, ∼0% RH, and five different temperatures (Table 1). The volume of the FLEC is 35 mL with other properties shown in Table S1 of the Supporting Information. 910

dx.doi.org/10.1021/es2035625 | Environ. Sci. Technol. 2012, 46, 909−915

Environmental Science & Technology

Article

difference using the ideal gas law (Vhigh = Vlow · Thigh/Tlow). Further, six-point calibration curves (R2 > 0.98) were developed for the analysis of each batch of samples, and the background DEHP in the sampling and analytical system was measured. The detection limit was 31 ng per adsorbent tube corresponding to 0.02 − 12 ng·m−3; this was determined as three times the standard deviation of the mean concentration of a low-concentration standard (n = 18) which resulted in an amount of 70 ng DEHP per adsorbent tube. Quality control using the internal standard D4-DEHP was carried out using one-concentration level DEHP control standards.22 In brief, a batch of stable control standards in sealed ampules was prepared and the mean concentration estimated from TD-GCMS analysis of 20 control standards before the FLEC experiments were carried out. One control standard was included in every batch of samples analyzed. None of those control standards were significantly different from the established mean concentration of the control standard. Turbomass software (Perkin-Elmer) was used to treat the chromatographic data and Microsoft Office Excel 2003 and 2010 were used as spreadsheets. Estimation of the Surface Concentration, Csurf, of DEHP Adsorbed to the Internal Surface of the FLECs. The interior surface of each FLEC, with an area of 0.018 m2, was rinsed with methanol after termination of the experiment to estimate the surface concentration of DEHP at steady state, Csurf, according to Clausen et al..10 The recovery of 102% ± 11% (N = 10) has previously been established.12 The surface of each FLEC was rinsed with 30−40 mL methanol using a glass pipet. A concentration step was necessary by injection of 30− 50 μL of the methanol solution into Tenax tubes, purging with nitrogen, and analyzing with TD-GC-MS as previously described.10,19 Computational Fluid Dynamics (CFD). CFD Method. The CFD model developed previously12 was employed for all the CFD simulations in the present paper and the model is summarized here. The cross section profile of the FLEC is shown in Figure S1 of the Supporting Information. Due to the symmetrical structure, one-quarter of the entire FLEC was modeled, as shown in Figure S2 of the Supporting Information. Under all experimental conditions, the air flow in the FLEC cavity was laminar.23 Employing FLUENT 6.3.26 (Fluent Inc., Lebanon, NH, U.S.), the flow field was solved with uniform air inlet velocity as boundary condition. To calculate the DEHP concentration in the simulation domain, a user-defined-scalar (UDS) function was introduced based on the advectiondiffusion equation:

∂(ujC) ∂C ∂ 2C + = Da 2 ∂t ∂xj ∂x j

Cs(t + Δt ) − Cs(t ) Δt Δt → 0 K steelCa(t + Δt ) − K steelCa(t ) = − lim Δt Δt → 0

Flux(t ) = − lim

= − K steelCa′(t )

(2) −2 −1

where Flux(t) is the mass flux, or DEHP loss rate (ug·m ·s ) in the air at a specific position adjacent to the upper internal stainless steel surface due to adsorption onto the surface. Cs(t) is the internal stainless steel surface concentration at the specific position and Ca(t) is the gas-phase concentration immediately adjacent to the position on the surface. Ksteel is the partition coefficient of DEHP between the FLEC chamber air and the internal stainless steel surface. This flux is the boundary condition for the upper internal stainless steel surface (adsorption surface), while the fluxes for all the other surfaces were assumed to be zero. As shown previously12 when the flow rate is low (450 mL·min−1), the chamber concentration at steady state is essential uniform and the steady-state outlet concentration, Csteady, is approximately equal to y0. Estimation of CFD Parameters. Two main parameters are required for CFD simulation of the DEHP concentration over time in the FLECs at different temperatures. These are y0 and the steel/air partition coefficient, Ksteel, at different temperatures. The steady-state concentrations, Csteady, at different temperatures (see Table 1) are used as approximations of y0 as discussed previously.12 Ksteel is assumed to be constant (a linear isotherm) over the entire concentration range and is calculated as Ksteel = Csurf/Csteady and shown in Table 1 for the different temperatures. Ksteel’s dependence on temperature is shown in Figure 1, where the lines fitted with nonlinear regression (SPSS

Figure 1. Ksteel (primary axis) and y0 (secondary axis) as a function of temperature. Closed circles are the partition coefficient of DEHP, Ksteel, between FLEC air steady-state concentration and measured surface concentration. Open circles are Ksteel values obtained by “fitting” CFD curves to FLEC concentrations versus time data by varying Ksteel. The open squares are y0 approximated by the steadystate concentrations, Csteady, of DEHP emitted from vinyl flooring in the FLEC. The red line is concentration predicted using that y0 (Csteady) = vapor pressure of pure DEHP (PT(DEHP)) at different temperatures.

(1)

where C is the DEHP gas-phase concentration, t is time, uj is the velocity component on the coordinate xj, and Da is the diffusivity of DEHP in air. For the emission simulation, the boundary condition for the inlet air was Cinlet = 0 and the DEHP concentration immediately adjacent to the vinyl flooring surface was a constant (y0). An instantaneously reversible equilibrium was assumed between the upper internal stainless steel surface and the air at the interface.14 Thus, the mass flux of DEHP at every location of the FLEC upper surface was given by the following:

statistical software, IBM) are based on a model for Ksteel dependence of temperature by Zhang et al.:24 Ksteel = P1 · T1/2 · exp(P2/T). For measured Ksteel: P1 = 7.53 × 10−20, and P2 = 1.41 × 104; for CFD “fitted” Ksteel: P1 = 6.94 × 10−1, and P2 = 911

dx.doi.org/10.1021/es2035625 | Environ. Sci. Technol. 2012, 46, 909−915

Environmental Science & Technology

Article

8.13 × 103. For the 23 °C case the mean of the values in Table 1 are used in the CFD simulations. In addition, diffusivity of DEHP in air, and viscosity and density of air as a function of temperature are shown in Table S2 in the Supporting Information. Vapor Pressure of DEHP at Different Temperatures. Vapor pressure of DEHP (PT(DEHP)) has been measured and deduced over many years using a variety of methods for temperatures from 17 to 230 °C25−33 and has resulted in remarkably consistent values. Assuming ΔH is constant over the temperature range (17 to 230 °C) and fitting the Clausius− Clapeyron relation, ln(PT(DEHP)) = −ΔH/R· (1/T) + C, where ΔH (J·mol−1) is the enthalpy of evaporation and R (8.314 J·K−1·mol−1) is the gas constant, gives ΔH/R = 12 891K corresponding to ΔH = 107 kJ·mol−1. Considering that ΔH may itself be a function of temperature, ΔH was assumed to be a constant between 17 and 48 °C which is close to the temperature range used in our measurements. The linear regression resulted in slope and intercept that are slightly different compared to the result for the full temperature range. Thus ΔH/R = 11 452 K corresponds to ΔH = 95 ± 1 kJ/mol and this value of ΔH was used to calculate PT(DEHP) at 23 °C, 35 °C, 47 °C, 55 °C, and 61 °C. The estimated ΔH = 95 ± 1 kJ·mol−1 agrees well with the value of 98 ± 2 kJ·mol−1 for evaporation of DEHP from a glass plate at 60−100 °C measured by thermogravimetry.17 The detailed analysis of vapor pressures from literature is shown in the Supporting Information (Figures S4 and S5). Modeling the DEHP Concentration versus Time in a Home at 23 and 35 °C. Emission of DEHP from vinyl flooring is subject to “external” control.14 This means that the emission rate is proportional to the convective mass-transfer coefficient, hm, and the difference between the DEHP concentration immediately adjacent to the test piece surface, y0, and the bulk air concentration. The bulk air concentration in turn depends on the partition coefficients for adsorption to indoor surfaces, K, and airborne particles, Kp. The threecompartment model used here was previously published34 and predicts the DEHP concentrations over time in a house with bathroom, kitchen, and main house. The floors in the bathroom and kitchen are fully covered and in the main house partly covered with vinyl flooring containing DEHP. In order to carry out the modeling we needed to estimate the values of hm, y0, K, and Kp at 23 and 35 °C, as described in detail in the Supporting Information (Table S3 and Figure S6).

Figure 2. Measured DEHP concentration in the FLEC at 23 °C and CFD simulation for measured (red) and “fitted” (blue) Ksteel values.

Figure 3. Measured DEHP concentration in the FLEC at 47 °C and CFD simulation for measured (red) and “fitted” (blue) Ksteel values.

duplicate sampling on Tenax TA and analysis by TD-GC-MS as previously discussed in the Supporting Information of Clausen et al.12 In addition, quality control as described in “Sampling and analysis of DEHP in FLEC air” in the Experimental Section did not reveal any problems with the analysis method. Measurements and CFD Simulations. All CFD simulations agree with the experimental results although they tend to overestimate concentrations as they approach steady-state. This is most pronounced at 47 °C as indicated by the “fitted” blue CFD graph in Figure 3. This deviation may be due to (1) incomplete extraction of DEHP from FLEC surfaces with methanol, so that Ksteel is underestimated or (2) a nonlinear adsorption isotherm for DEHP in the FLECs. In our previous study, the recovery of DEHP from the internal FLEC surface was 102% ± 11% (N = 10) obtained by sprinkling the surface with 250 μL methanol solution of DEHP (70 μg), evaporation of methanol and finally rinsing with pure methanol and analysis.12 In addition, we have some preliminary methanol extraction data that suggest a nonlinear adsorption isotherm for DEHP in the FLEC. However, it has been shown for a stainless steel chamber that the adsorption isotherm is linear and that recovery from solvent rinsing of the stainless steel is much lower than thermal desorption.35,36 Furthermore, we have observed large variations of measured surface concentrations of DEHP (Csurf) in the FLECs, previously,12 and in this study (see Table 1). However, this variation is of no significance for the relative Ksteel values at different temperatures since Csteady covers a much wider range than Csurf and it does not explain the apparently strong underestimation of Ksteel (see Figure 1). A possible explanation for these seemingly inconsistent findings is

■

RESULTS AND DISCUSSION The concentration versus time data for emission of DEHP from vinyl flooring in the FLECs at 23 and 47 °C including the CFD simulated concentrations for measured and CFD “fitted” Ksteel values are shown in Figures 2 and 3. Data for the other temperatures are presented in the Supporting Information (Figures S7−S9). The measured steady-state concentrations, Csteady, which are used as an approximation of y0, increase strongly with increasing temperature as shown Table 1 and Figure 1. The 38 °C increase in temperature (23−61 °C) resulted in a 211-fold increase in Csteady. The background concentrations of DEHP in the FLECs before the experiment were all below the detection limit. However, it cannot be excluded that the relatively large variation in concentrations at the 23 °C experiments may be due to variations in the background concentrations of DEHP. The large variability for some DEHP concentrations is what may be expected for 912

dx.doi.org/10.1021/es2035625 | Environ. Sci. Technol. 2012, 46, 909−915

Environmental Science & Technology

Article

distribution. The vinyl flooring as emission source is characterized by a constant DEHP gas-phase concentration, y0, immediately adjacent to the emission surface which drives diffusion of DEHP from the emission surface into the FLEC cavity.12 Because the flow within the FLEC cavity is laminar,23 diffusion in the air next to the surface of the vinyl flooring determines how fast DEHP can be transferred from the emission surface into the bulk air in the vertical direction. Gasphase DEHP is then strongly adsorbed onto the interior surface. When temperature increases, y0, which directly correlates with the vapor pressure of pure liquid DEHP (Figure 4), increases greatly and the diffusivity of DEHP in air also increases (Table S2 of the Supporting Information), leading to a higher emission rate. As a result, the steady-state concentration increases with temperature. The partition coefficient between the stainless steel surface and air decreases with temperature (the sorption effect decreases, Figure 1) and this allows the gas-phase concentration to reach steady state more rapidly at elevated temperature (compare Figures 2, 3 and S7−S9 of the Supporting Information). Modeling the DEHP Concentration versus Time in a Home at 23 and 35 °C. Studies including field measurements of temperatures in buildings and other indoor environments are scarce in peer-reviewed literature.44 However, we were able to find a few studies45−48 representing a wide range of publication times and locations and extracted approximate data on temperatures in buildings as described in the Supporting Information. These studies showed average temperatures in the range 16−24 °C and maximum temperatures in the range 29− 34 °C. In cars maximum temperatures may reach 58−67 °C.49 Thus, these studies show the relevance of modeling emissions at 23 and 35 °C. Furthermore, incident sunlight through windows may increase DEHP emission from plasticized vinyl products substantially by heating up the surface. The modeled gas and particle phase DEHP concentrations versus time in a home at 23 and 35 °C are shown in Figure 5 and Table 2. In the FLECs, which contain negligible amounts of airborne particles, the ratio between the pure gas-phase Csteady at 35 and 23 °C is about 10 whereas the ratio of the gas-phase DEHP concentrations in the modeled cases at 600 days is about 24. However, a larger fraction is in the gas-phase at 35 °C. The ratio of the sums of the gas-phase and the particle-phase is about nine, i.e., close to the ratio in the FLECs. The gas-phase concentration in a room is strongly impacted by airborne particles due to sorption of a large fraction of DEHP onto the particles and thus eliminated from the room by the particlephase. Considering the mass balance at steady-state in a wellmixed volume, the steady-state concentration can be derived as follows:50

that some DEHP was absorbed slowly into the micropores on the stainless steel wall and could not be easily extracted by rinsing with methanol. During the chamber tests, which took several days to months, DEHP may have had sufficient time to enter the micropores so that rapid methanol extraction was less effective than thermal desorption. The inconsistency of results for adsorption isotherms and surface concentrations need to be tested by in a thorough investigation of adsorption isotherms and recovery of DEHP, which is beyond the scope of this work. Another problem, which is discussed in the Supporting Information, is that the measured Ksteel for DEHP in the FLECs appear to decrease over time from one study to another. However, Ksteel is only of importance when the system is not at steady state and choosing the measured Ksteel values instead of the “fitted” values will not change the conclusions of this work. Vapor Pressure of Pure DEHP and DEHP in Vinyl Flooring. Ekelund et al.11 observed that DEHP in PVC evaporated at the same rate as pure DEHP and they suggested that this was due to a thin film of DEHP present on the PVC surface. We found that the concentration immediately adjacent to the vinyl flooring y0 (approximated by Csteady) is equal to the vapor pressure of pure DEHP at the five different temperatures (Figure 4) and in this way confirmed the observation by

Figure 4. Plot of y0 versus vapor pressure of pure DEHP expressed in concentration units at different temperatures. y0 is approximated by the steady-state concentrations in the FLECs, Csteady. The regression line is forced through 0,0 and shows that they are approximately equal. The vapor pressures of DEHP are calculated from the regression line in Figure S5 of the Supporting Information.

Ekelund et al.11 The linear relationship between Csteady and PT(DEHP) means that the DEHP emission from PVC can be predicted in a fast and accurate way since Csteady at 61 °C has low uncertainty and is obtained in a short test period (see Csteady time interval in Table 1). A German standard for the determination of SVOCs from polymer materials already recommend elevated test chamber temperatures.37,38 The reason for the high vapor pressure of DEHP in PVC may stem from the fact that SVOCs in certain matrices exist in a thermodynamically separated phase with liquid-like properties.39−41 We note that Raoult’s Law does not apply for polymer solutions such as the DEHP/vinyl flooring system due to the extreme difference in molecular size of the polymer and DEHP 42,43 and the corresponding large deviations from ideal behavior. The Emission Mechanism of DEHP from Vinyl Flooring. Because the concentration in the FLEC cavity is not uniform, especially before steady state is reached,12 we use a CFD model that considers the nonuniform concentration

y=

hmy0 A hmA + (1 + K p·TSP) ·Q

(3)

where y is the gas-phase concentration, A is the emission surface area, TSP is the concentration of total suspended particles, and Q is the air flow rate. At higher temperature, hm and y0 are larger while Kp decreases dramatically. When particles are present, y increases with temperature for two reasons: (1) higher emission rates driven by higher hm and higher y0; (2) less DEHP sorption onto particles due to lower Kp. Therefore, the increasing effect of y at higher temperature in the FLEC is not as large as that in a room with airborne particles. Meanwhile, K (partition coefficient between indoor 913

dx.doi.org/10.1021/es2035625 | Environ. Sci. Technol. 2012, 46, 909−915

Environmental Science & Technology

■

Article

ASSOCIATED CONTENT

S Supporting Information *

Further details on FLEC properties, CFD modeling, modeling of DEHP concentrations in a home at 23 and 35 °C, change of Ksteel in the FLEC over time, evaluation of literature data on vapor pressure of DEHP at different temperatures, literature on temperatures in building and cars, and concentration versus time data for emission of DEHP from vinyl flooring in the FLEC at 35, 55, and 61°C. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +45 39 16 52 73; fax: +45 39 16 52 01; e-mail: pac@ nrcwe.dk.

■

ACKNOWLEDGMENTS We thank Real Dania for financial support and Larisa Jelezneacova for technical assistance.

■ Figure 5. The modeled DEHP concentration versus time in a home at (a) 23 °C and (b) 35 °C. Note that the gas-phase and particle-phase have differently scaled axes. The model parameters are shown in Table S3 and Figure S6 in the Supporting Information.

Table 2. Estimated Partition Coefficients Used to Predict the DEHP Concentration in a Home at 23 and 35 °C surface

partition coefficient, K @ 23 °C18

partition coefficient, K @ 35 °C

2500

375

1700 3800

255 570

furniture, wood floor, wall and ceiling (m) carpet (m) glass and tiles (μg·m−2)·(μg·m−3)−n a a

REFERENCES

(1) Rudel, R. A.; Perovich, L. J. Endocrine disrupting chemicals in indoor and outdoor air. Atmos. Environ. 2009, 43, 170−181. (2) Clausen, P. A.; Bille, R. L. L.; Nilsson, T.; Hansen, V.; Svensmark, B.; Bøwadt, S. Simultaneous extraction of di(2-ethylhexyl)phthalate and non-ionic surfactants from house dust. Concentrations in floor dust from 15 Danish schools. J. Chromatogr. A 2003, 986, 179−190. (3) Rudel, R. A.; Camann, D. E.; Spengler, J. D.; Korn, L. R.; Brody, J. G. Phthalates, alkylphenols, pesticides, polybrominated diphenyl ethers, and other endocrine-disrupting compounds in indoor air and dust. Environ. Sci. Technol. 2003, 37, 4543−4553. (4) Øie, L.; Hersoug, L.-G.; Madsen, J. Ø. Residential exposure to plasticizers and its possible role in the pathogenesis of asthma. Environ. Health Perspect. 1997, 105, 972−978. (5) Bornehag, C.-G.; Sundell, J.; Weschler, C. J.; Sigsgaard, T.; Lundgren, B.; Hasselgren, M.; Hägerhed-Engman, L. The association between asthma and allergic symptoms in children and phthalates inhouse dust: a nested case-control study. Environ. Health Perspect. 2004, 112, 1393−1397. (6) Kolarik, B.; Naydenov, K.; Larsson, M.; Bornehag, C.-G.; Sundell, J. The association between phthalates in dust and allergic diseases among Bulgarian children. Environ. Health Perspect. 2008, 116, 98− 103. (7) Bornehag, C. G.; Nanberg, E. Phthalate exposure and asthma in children. Int. J. Androl. 2010, 33, 333−345. (8) Jaakkola, J. J. K.; Knight, T. L. The role of exposure to phthalates from polyvinyl chloride products in the development of asthma and allergies: A systematic review and meta-analysis. Environ. Health Perspect. 2008, 116, 845−853. (9) Nielsen, G. D.; Larsen, S. T.; Olsen, O.; Løvik, M.; Poulsen, L. K.; Glue, C.; Wolkoff, P. Do indoor chemicals promote development of airway allergy? Indoor Air 2007, 17, 236−255. (10) Clausen, P. A.; Hansen, V.; Gunnarsen, L.; Afshari, A.; Wolkoff, P. Emission of di(2-ethylhexyl) phthalate from PVC into air and dust. Emission and sorption experiments. Environ. Sci. Technol. 2004, 38, 2531−2537. (11) Ekelund, M.; Azhdar, B.; Gedde, U. W. Evaporative loss kinetics of di(2-ethylhexyl)phthalate (DEHP) from pristine DEHP and plasticized PVC. Polym. Degrad. Stab. 2010, 95, 1789−1793. (12) Clausen, P. A.; Liu, Z.; Xu, Y.; Kofoed-Sørensen, V.; Little, J. C. Influence of air flow rate on emission of DEHP from vinyl flooring in the emission cell FLEC: Measurements and CFD simulation. Atmos. Environ. 2010, 44, 2760−2766. (13) Clausen, P. A.; Xu, Y.; Kofoed-Sørensen, V.; Little, J. C.; Wolkoff, P. The influence of humidity on the emission of di-(2-

Isotherm exponent n = 1.5.

surfaces and air) decreases with an increase in temperature, leading to less migration of DEHP onto interior surfaces from air. In other words, at elevated temperature, it takes a shorter time to fill the sorption reservoir with DEHP because a larger fraction of DEHP is in the gas-phase compared to the sorbed. It therefore takes less time to reach steady state at higher temperature. The conclusions of this work should be applicable to emission and transport of other phthalates and other SVOCs contained in solid materials. The overall impact of higher temperature, whether due to increased y0 or due to decreased K and Kp, is accelerated emissions and elevated gas-phase concentrations. As a result, significantly higher human exposure via inhalation of air would occur while the total exposure (including inhalation of air, inhalation of airborne particles, ingestion of dusts, and dermal absorption) depends on specific chemical properties (e.g., Kp) and environmental factors (e.g., TSP). 914

dx.doi.org/10.1021/es2035625 | Environ. Sci. Technol. 2012, 46, 909−915

Environmental Science & Technology

Article

ethylhexyl) phthalate (DEHP) from vinyl flooring in the emission cell “FLEC”. Atmos. Environ. 2007, 41, 3217−3224. (14) Xu, Y.; Little, J. C. Predicting emissions of SVOCs from polymeric materials and their interaction with airborne particles. Environ. Sci. Technol. 2006, 40, 456−461. (15) Kim, J. H.; Kim, S. H.; Lee, C. H.; Nah, J.; Hahn, A. DEHP migration behavior from excessively plasticized PVC sheets. Bull. Korean Chem. Soc. 2003, 24, 345−349. (16) Fujii, M.; Shinohara, N.; Lim, A.; Otake, T.; Kumagai, K.; Yanagisawa, Y. A study on emission of phthalate esters from plastic materials using a passive flux sampler. Atmos. Environ. 2003, 37, 5495− 5504. (17) Ekelund, M.; Azhdar, B.; Hedenqvist, M. S.; Gedde, U. W. Long-term performance of poly(vinyl chloride) cables, Part 2: Migration of plasticizer. Polym. Degrad. Stab. 2008, 93, 1704−1710. (18) Xu, Y.; Hubal, E. C.; Clausen, P. A.; Little, J. C. Predicting Residential Exposure to Phthalate Plasticizer Emitted from Vinyl Flooring: A Mechanistic Analysis. Environ. Sci. Technol. 2009, 43, 2347−2380. (19) Kofoed-Sørensen, V.; Clausen, P. A. Preconcentration and analysis of phthalate esters in house dust extracts - The advantages of using thermal desorption and gas chromatography. G. I. T. Lab. J. 5 2004, 8, 34−35. (20) Wolkoff, P. An emission cell for measurement of volatile organic compounds emitted from building materials for indoor useThe field and laboratory emission cell FLEC. Gefahrstoffe - Reinhalt. Luft 1996, 56, 151−157. (21) Wolkoff, P.; Clausen, P. A.; Nielsen, P. A.; Gustafsson, H.; Jonsson, B.; Rasmusen, E. Field and Laboratory Emission Cell: FLEC. In: IAQ ’91 Healthy Buildings; Geshwiler, M., Montgomery, L., Moran, M., Ed.; American Society of Heating, Refrigerating and AirConditioning Engineers (ASHRAE): Atlanta, 1991; pp 160−165. (22) Clausen, P. A. Kofoed-Sørensen, V. Sampling and analysis of SVOCs and POMs in indoor air. In Organic Indoor Air Pollutants; Salthammer, T. Uhde, E., Eds.; Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, 2009; pp 19. (23) Zhu, Q.; Kato, S.; Murakami, S.; Ito, K. 3D-CFD analysis of diffusion and emission of VOCs in a FLEC cavity. Indoor Air 2007, 17, 178−188. (24) Zhang, Y.; Luo, X.; Wang, X.; Qian, K.; Zhao, R. Influence of temperature on formaldehyde emission parameters of dry building materials. Atmos. Environ. 2007, 41, 3203−3216. (25) Small, P. A.; Small, K. W.; Cowley, P. The vapour pressures of some high boiling esters. Trans. Faraday Soc. 1948, 44, 810−816. (26) Werner, A. C. Vapor pressures of phthalate esters. Ind. Eng. Chem. Res. 1952, 44, 2736−2740. (27) Quackenbos, H. M. Plasticizers in vinyl chloride resins. Ind. Eng. Chem. Res. 1954, 46, 1335−1349. (28) Ray, A. K.; Davis, E. J.; Ravindran, P. Determination of ultra-low vapor pressures by submicron droplet evaporation. J. Chem. Phys. 1979, 71, 582−587. (29) Dobbs, A. J.; Cull, M. R. Volatilisation of chemicalsRelative loss rates and the estimation of vapour pressures. Environ. Pollut., Ser. B 1982, 3, 289−298. (30) Hinckley, D. A.; Bidleman, T. F.; Foreman, W. T.; Tuschall, J. R. Determination of vapor pressures for nonpolar and semipolar organic compounds from gas chromatograhic retention data. J. Chem. Eng. Data 1990, 35, 232−237. (31) Tang, I. N.; Munkelwitz, H. R. Determination of vapor pressure from droplet evaporation kinetics. J. Colloid Interface Sci. 1991, 141, 109−118. (32) Price, D. M. Vapor pressure determination by thermogravimetry. Thermochim. Acta 2001, 367−368, 253−262. (33) Clausen, P. A.; Hansen, V.; Gunnarsen, L.; Afshari, A.; Wolkoff, P. Emission of Phthalates from PVC flooring in two very different test chambers. In: Indoor Air 2002Proceedings of the 9th International Conference on Indoor Air Quality and Climate; Levin, H., Ed.; Indoor Air 2002: Monterey, California; USA, 2002; Vol. 2, pp 932−937.

(34) Xu, Y.; Cohen Hubal, E. A.; Little, J. C. Predicting residential exposure to phthalate plasticizer emitted from vinyl flooring: sensitivity, uncertainty, and implications for biomonitoring. Environ. Health Perspect. 2010, 118, 253−258. (35) Liu, Z.; Xu, Y.; Little, J. C. Characterizing emissions of di-2ethylhexyl phthalate from vinyl flooring in a specially-designed chamber. In Indoor Air 2011 - Proceedings of the 12th International Conference on Indoor Air Quality and Climate; Indoor Air 2011: Austin, Texas, 2011. (36) Xu, Y.; Park, J.; Kofoed-Sørensen, V.; Clausen, P. A.; Little, J. C. Characterizing emission of phthalate plasticizer from vinyl flooring in a specially-designed chamber. Epidemiology 2008, 19, S294−S295. (37) Anon. Determination of organic substances as emitted from automotive interior products using a 1 m3 test cabinet. 2000;VDA 276: (38) Schripp, T.; Nactwey, B.; Toelke, J.; Salthammer, T.; Uhde, E.; Wensing, M.; Bahadir, M. A microscale device for measuring emissions from materials for indoor use. Anal. Bioanal. Chem. 2007, 387, 1907− 1919. (39) Oja, Vahur, Suuberg, Eric M. Vapor liquid equilibrium in polycyclic aromatic compound mixtures and in coal tars. In Heavy Hydrocarbon Resources; American Chemical Society: Washington, DC, USA, 2005; pp 113. (40) Burks, G. A.; Harmon, T. C. Volatilization of solid-phase polycyclic aromatic hydrocarbons from model mixtures and lampblackcontaminated soils. J. Chem. Eng. Data 2001, 46, 944−949. (41) Goldfarb, J. L.; Suuberg, E. M. Raoult’s Law and its application to sublimation vapor pressures of mixtures of polycyclic aromatic hydrocarbons. Environ. Eng. Sci. 2008, 25, 1429−1438. (42) Hawkes S. J. Raoult’s Law is a deception. J. Chem. Educ. 1995, 72, 204-null. (43) Nicholson, J. W. The Chemistry of Polymers; The Royal Society of Chemistry: Cambridge, 2006. (44) Humphreys, M. A.; Nicol, J. F.; Raja, I. A. Field studies of indoor thermal comfort and the progress of the adaptive approach. Adv. Building Energy Res. 2007, 1, 55−88. (45) Hunt, D. R. G.; Gidman, M. I. A national field survey of house temperatures*. Build. Environ. 1982, 17, 107−124. (46) Nicol, F.; Roaf, S. Pioneering new indoor temperature standards: the Pakistan project. Energy Build. 1996, 23, 169−174. (47) Raja, I. A.; Nicol, J. F.; McCartney, K. J.; Humphreys, M. A. Thermal comfort: Use of controls in naturally ventilated buildings. Energy Build. 2001, 33, 235−244. (48) Rousseau, M.; Manning, M.; Said, M. N.; Cornick, S. M.; Swinton, M. C. Characterization of indoor hygrothermal conditions in houses in different northern climates. In: Thermal Performance of Exterior Envelopes of Whole Buildings X International Conference; Thermal Performance of Exterior Envelopes of Whole Buildings X International Conference; Clearwater, Florida, 2007; vol. pp. 1-14. (49) McLaren, C.; Null, J.; Quinn, J. Heat stress from enclosed vehicles: moderate ambient temperatures cause significant temperature rise in enclosed vehicles. Pediatrics 2005, 116, e109−e112. (50) Little, J. C.; Weschler, C. J.; Nazaroff, W. W.; Liu, Z.; Cohen Hubal, E. A. Semivolatile organic compounds in the indoor environmentToward rapidly characterizing exposure and risk (In preparation).

915

dx.doi.org/10.1021/es2035625 | Environ. Sci. Technol. 2012, 46, 909−915