gas partition coefficients for seventy-two semi-volatile organic compounds in indoor environment

Chemosphere 153 (2016) 212e219

Contents lists available at ScienceDirect

Chemosphere journal homepage: www.elsevier.com/locate/chemosphere

Distributions of the particle/gas and dust/gas partition coefficients for seventy-two semi-volatile organic compounds in indoor environment Wenjuan Wei a, *, Corinne Mandin a, b, c, Olivier Blanchard d, b, Fabien Mercier d, b, c, Maud Pelletier d, b, Barbara Le Bot d, b, c, Philippe Glorennec d, b, Olivier Ramalho a a University of Paris-Est, Scientific and Technical Center for Building (CSTB), Health and Comfort Department, French Indoor Air Quality Observatory (OQAI), 84 Avenue Jean Jaur
es, Champs Sur Marne, 77447 Marne la Vall ee Cedex 2, France b INSERM-U1085, Irset-Research Institute for Environmental and Occupational Health, Rennes, France c LERES-Environment and Health Research Laboratory (Irset and EHESP Technologic Platform), Rennes, France d EHESP-School of Public Health, Sorbonne Paris Cit e, Rennes, France

h i g h l i g h t s
38 empirical equations to calculate Kp and Kd from p0 and Koa were summarized. L
A reference distribution of log10 Kp for 72 SVOCs was determined.
The log10 Kp values were normally distributed for 27 SVOCs.
The distribution reduces the bias in choosing a specific equation to calculate Kp.
The ratio between Kp and Kd was determined using linear regression.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 21 January 2016 Received in revised form 3 March 2016 Accepted 3 March 2016

Particle/gas and dust/gas partition coefficients (Kp and Kd) are two key parameters that address the partitioning of semi-volatile organic compounds (SVOCs) between gas-phase, airborne particles, and settled dust in indoor environment. A number of empirical equations to calculate the values of Kp and Kd have been reported in the literature. Therefore, the difficulty lies in the selection of a specific empirical equation in a given situation. In this study, we retrieved from the literature 38 empirical equations for calculating Kp and Kd values from the SVOC saturation vapor pressure and octanol/air partition coefficient. These values were calculated for 72 SVOCs: 9 phthalates, 9 polybrominated diphenyl ethers (PBDEs), 11 polychlorinated biphenyls (PCBs), 22 biocides, 14 polycyclic aromatic hydrocarbons (PAHs), 3 alkylphenols, 2 synthetic musks, tributylphosphate, and bisphenol A. The mean and median values of log10 Kp or log10 Kd for most SVOCs were of the same order of magnitude. The distribution of log10 Kp values was fitted to either a normal distribution (for 27 SVOCs) or a log-normal distribution (for 45 SVOCs). This work provides a reference distribution of the log10 Kp for 72 SVOCs, and its use may reduce the bias associated with the selection of a specific value or equation. © 2016 Elsevier Ltd. All rights reserved.

Handling Editor: Ralf Ebinghaus Keywords: Gas-phase Airborne particles Partitioning Equilibrium Indoor air quality

1. Introduction The partitioning equilibrium of semi-volatile organic compounds (SVOCs) between gas-phase and airborne particles, and between gas-phase and settled dust in indoor environments can be

^timent (CSTB), * Corresponding author. Centre Scientifique et Technique du Ba -Confort e Observatoire de la Qualite  de l'Air Inte rieur (OQAI), 84 Direction Sante
s, Champs Sur Marne, 77447 Marne la Valle e Cedex 2, France. Avenue Jean Jaure E-mail address: [email protected] (W. Wei). http://dx.doi.org/10.1016/j.chemosphere.2016.03.007 0045-6535/© 2016 Elsevier Ltd. All rights reserved.

described by the particle/gas partition coefficient (Kp) and the dust/ gas partition coefficient (Kd), respectively (Weschler et al., 2008). At steady state, if Kp or Kd is known, unknown SVOC concentrations in the gas-phase can be calculated based on their measured concentrations in airborne particles or settled dust, respectively (Weschler and Nazaroff, 2010). The Kp and Kd values can be obtained by measuring SVOC concentrations of gas-phase, airborne particles, and settled dust under the assumption of a partitioning equilibrium of SVOCs within these three phases. For example, Benning et al. (2013) used vinyl flooring

W. Wei et al. / Chemosphere 153 (2016) 212e219

placed inside a stainless steel environmental chamber as a di-(2ethylhexyl)-phthalate (DEHP) source. They measured the DEHP concentrations in the gas-phase and airborne particles and calculated the particle/gas partition coefficient. The Kp and Kd values over a wide range of SVOCs are more frequently calculated using empirical equations that include the SVOC saturation vapor pressure (p0L ) or the octanol/air partition coefficient (Koa). Though a large amount of work is available, two practical problems concerning the determination of Kp and Kd for SVOCs remain unsolved. First, a number of equations exist for calculating Kp and Kd values. Theoretical relationships between Kp and p0L and between Kp and Koa were determined by Pankow (1994) and Finizio et al. (1997), respectively. However, some physical parameters in the theoretical equations, e.g., the activity coefficient of the absorbing SVOCs in airborne particles and the fraction of the airborne particles that is organic matter, have not been fully studied for most SVOCs and their values remain assumed in these calculations (Harner, 1998). Empirical equations, which are determined

213

based on the regression of experimental data for various compounds in a number of studies, differ in their regression constants. Finizio et al. (1997) critically reviewed 12 empirical equations that include Kp and p0L and proposed equations that include Koa as a sufficient descriptor to predict Kp for a number of SVOCs. Following their research study, more equations that include Koa were proposed by Harner (1998), Shoeib et al. (2005), and Weschler and Nazaroff (2010). Salthammer and Schripp (2015) performed a quantitative error analysis of the Junge-Pankow equation based on parameter uncertainties, but the error in other equations remains unverified. As a result, users may encounter difficulties in the selection of an equation for the prediction of Kp, Kd and the unknown concentration of SVOCs in a given phase. Second, the relationship between Kp and Kd has not been clearly addressed. Consequently, the accuracy of the gas-phase SVOC concentrations calculated using Kp or Kd remains unverified. To solve the problems, this work aims to provide reference distributions of log10 Kp and log10 Kd for a number of SVOCs with

Table 1 Empirical equations to calculate log10 Kp (m3 mg1) and log10 Kd (m3 g1) at 25  C. Empirical equation log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼ log10 Kp ¼

0:860 log10 p0L  4:67 log10 p0L  5:47 0:694 log10 p0L  4:61 0:81 log10 p0L  5:31 1:04 log10 p0L  5:95 0:76 log10 p0L  5:10 0:88 log10 p0L  5:38 0:61 log10 p0L  4:26 0:726 log10 p0L  5:18 0:95 log10 p0L  5:86 0:61 log10 p0L  4:74 0:69 log10 p0L  5:06 0:92 log10 p0L  5:63 0:8698 log10 p0L  4:7707 0:6027 log10 p0L  5:1661 0:715 log10 p0L  5:141 0:745 log10 p0L  4:666 0:65 log10 p0L  4:04 0:75 log10 p0L  4:35 0:86 log10 p0L  5:66 0:76 log10 p0L  4:57 0:51 log10 p0L  4:72 0:60 log10 p0L  4:93 0:58 log10 p0L  4:56 0:58 log10 p0L  4:72

Studied compounds

Reference

PAHs

Naumova et al., 2003

PAHs

Cotham and Bidleman, 1995; Finizio et al., 1997

PAHs

Cotham and Bidleman, 1995; Finizio et al., 1997

PAHs

Ngabe and Bidleman, 1992; Finizio et al., 1997

PAHs

Yamasaki et al., 1982; Finizio et al., 1997

PAHs

Foreman and Bidleman, 1990; Finizio et al., 1997

PAHs

Ligocki and Pankow, 1989; Finizio et al., 1997

PAHs

Baker and Eisenreich, 1990; Finizio et al., 1997

PCBs

Cotham and Bidleman, 1995; Finizio et al., 1997

PCBs

Foreman and Bidleman, 1990; Finizio et al., 1997

PCBs and organochlorine pesticides

Kaupp and Umlauf, 1992; Finizio et al., 1997

PCBs and organochlorine pesticides

Hoff et al., 1996; Finizio et al., 1997

Organochlorine pesticides

Foreman and Bidleman, 1990; Finizio et al., 1997

PAHs and polychlorinated naphthalenes (PCNs)

Kaupp and McLachlan, 1999

PCBs and organochlorine pesticides

Kaupp and McLachlan, 1999

PCBs

Harner, 1998

PAHs

Harner, 1998

PAHs

He and Balasubramanian, 2009

PAHs

He and Balasubramanian, 2009

PAHs

He and Balasubramanian, 2009

PAHs

He and Balasubramanian, 2009

PCBs

He and Balasubramanian, 2009

PCBs

He and Balasubramanian, 2009

PCBs

He and Balasubramanian, 2009

PCBs

He and Balasubramanian, 2009

PAHs PCBs and organochlorine pesticides PAHs PCBs and organochlorine pesticides SVOCs

Finizio et al., 1997; Weschler et al., 2008 Finizio et al., 1997 Finizio et al., 1997 Finizio et al., 1997 Weschler and Nazaroff, 2010

log10 Kp ¼ 0.6368 log10 Koa  8.9111 log10 Kp ¼ 0.654 log10 Koa  9.183 log10 Kp ¼ 0.735 log10 Koa  9.947 log10 Kp ¼ 0.829 log10 Koa  10.263 log10 Kp ¼ log10 Koa þ log10 fom p  11:91   b om d log10 Kd ¼ log10 Koa þ log10 0:411f rd   log10 Kd ¼ log10 Koa þ log10 fomr d c

PCBs, organochlorine pesticides, PAHs, and PCNs PCBs PCNs PAHs SVOCs Perfluorinated alkyl sulfonamides

Kaupp and McLachlan, 1999 Harner, 1998 Harner, 1998 Harner, 1998 Harner, 1998; Shoeib et al., 2005; Weschler et al., 2008

SVOCs

Weschler and Nazaroff, 2010

log10 Kd ¼ 0.86 log10 Koa  6.09

SVOCs

Weschler and Nazaroff, 2010

log10 Kp ¼ log10 Kp ¼ log10 Koa þ log10(1.88  1012) log10 Kp ¼ log10 Koa þ log10(1.5  1012) log10 Kp ¼ 0.79 log10 Koa  10.01 log10 Kp ¼ 0.55 log10 Koa  8.23 ! log10 Kp ¼ log10 Koa þ log10

fom p rp

a

d

a b c

fom fom fom

p d d

3

¼ 0:4, rp ¼ 1  mg m (Weschler and Nazaroff, 2010). ¼ 0:2 (Weschler and Nazaroff, 2010), rd ¼ 1  106 g m3 (Weschler et al., 2008). 6 ¼ 0:2, rd ¼ 2  10 g m3 (Weschler and Nazaroff, 2010). 1012

214

W. Wei et al. / Chemosphere 153 (2016) 212e219

Table 2 Studies providing log10 p0L or log10 Koa values for SVOCs. Reference

Weschler et al., 2008 Weschler, 2003 Weschler and Nazaroff, 2008

Studied compounds log10 p0L

log10 Koa

Phthalates

Phthalates

DEHP and PAHs e Phthalates, PBDEs, PCBs, biocides (pyrethroid, organochlorine and Phthalates, PBDEs, PCBs, biocides (pyrethroid, organochlorine and organophosphorous pesticides, and triclosan etc.), PAHs, bisphenol A, and organophosphorous pesticides, and triclosan etc.), PAHs, bisphenol A, and galaxolide galaxolide Finizio et al., 1997 PCBs, organochlorine pesticides, and PAHs PCBs, organochlorine pesticides, and PAHs Harner and DDT DDT Mackay, 1995 Shoeib and Organochlorine pesticides Organochlorine pesticides Harner, 2002 Weschler and e Phthalates, PBDEs, PCBs, biocides (pyrethroid, organochlorine and Nazaroff, 2010 organophosphorous pesticides), PAHs, and bisphenol A Li et al., 2006 e PBDEs, PCBs, and PAHs Chen et al., 2003 e PBDEs Odabasi and e Chlorpyrifos Cetin, 2012 Odabasi et al., e PAHs 2006

the objectives of reducing the bias associated with the selection of a specific equation and allowing the prediction of the distribution of unknown SVOC concentrations in a given phase using a MonteCarlo simulation method in future studies. Thus, the specific objectives of this study are (1) to use the entire set of published equations for calculating log10 Kp and log10 Kd to determine distributions of the corresponding values for a number of target SVOCs and (2) to investigate the empirical relationship between Kp and Kd for these SVOCs.

2.2. Target SVOCs The distribution values of Kp or Kd were determined for 72 target indoor SVOCs measured in French national surveys (Mandin et al., 2014a, b): 9 phthalates, 9 polybrominated diphenyl ethers (PBDEs), 11 polychlorinated biphenyls (PCBs), 22 biocides (pyrethroid, organochlorine and organophosphorous pesticides, and triclosan), 14 polycyclic aromatic hydrocarbons (PAHs), 3 alkylphenols, 2 synthetic musks, tributylphosphate (TBP), and bisphenol A (BPA).

2. Materials and methods

2.3. Vapor pressures and octanol/air partition coefficients

2.1. Selection of empirical equations to calculate Kp and Kd

Kp and Kd were calculated from p0L and Koa. Two approaches were used to retrieve p0L and Koa. First, the values of log10 p0L and log10 Koa at 25  C for the target SVOCs were searched in the literature. Second, log10 p0L and log10 Koa for the target SVOCs were obtained at 25  C using the EPI Suite v4.11 software (Environmental Protection Agency, 2015). The Koa values were retrieved from both estimates based on octanol/water partition coefficients (using the KOAWIN calculator) and from experimental results presented in the EPI Suite software. The p0L values were estimated using the MPBPWIN calculator in the EPI Suite software.

The coefficient values of Kp (m3 mg1) and Kd (m3 mg1) are defined by the following equations (Weschler and Nazaroff, 2010):

Kp ¼

F Cg TSP

(1)

Kd ¼

Xd Cg

(2)

where F (mg m3), Cg (mg m3), and Xd (mg mg1) are SVOC equilibrium concentrations in an indoor environment for airborne particles, gas-phase, and settled dust, respectively, and TSP (mg m3) is the equilibrium concentration of the total suspended particles in indoor air. Based on the theoretical relationships between Kp and p0L and between Kp and Koa, empirical equations for calculating Kp and Kd were developed in previous studies by the regression of Kp, Kd, p0L , and Koa experimental data. In the present study, the terms “particle gas partition coefficient”, “dust gas partition coefficient”, “octanol”, and “vapor pressure” were used in the “Science Direct” and “Google Scholar” search engines, irrespective of publication year. Specific papers that developed original empirical equations to calculate Kp or Kd from p0L or Koa or that presented existing empirical equations with new parameters were retrieved. The original relationships developed for a specific chemical family of SVOCs were assumed to be applicable to all SVOCs.

2.4. Calculation of Kp and Kd, and the relationship between these two parameters The values of Kp and Kd were calculated using empirical equations retrieved from the literature. All available log10 p0L and log10 Koa values were used as input parameters. Each calculated Kp or Kd was assumed to have the same weight, which accounted for the same credit for further analysis. The calculated dataset of Kp and Kd for each SVOC was statistically analyzed, and the distribution was determined. The values of Kp and Kd were expected to be proportional to the octanol/air partition coefficient (Koa) and to the fraction of airborne particles or settled dust that is organic matter (fom p or fom d ), respectively. These values were also expected to be inversely proportional to the density of particles and of settled dust (rp or rd ) (Weschler and Nazaroff, 2010). Therefore, the expected relationship between Kp and Kd can be derived using the following equation:

W. Wei et al. / Chemosphere 153 (2016) 212e219

215

Table 3 Percentiles of log10 Kp and log10 Kd for SVOCs at 25  C. Compound

Phthalates DEHP DiNP BBP DiBP DBP DEP DMP DMEP DOP PBDEs 28 47 85 99 100 119 153 154 209 28 PCBs 31 52 77 101 105 118 126 138 153 180 Biocides TRIC 4,40 -DDT 4,40 -DDE g-HCH a-HCH PMET CYFL CYPE DMET OXAD DIEL a-ENDO CHLO DICH METO DIAZ TCHL ALDR ATRA ENDR HEPT CCHL PAHs BPER BKFL BBFL CHRY FLUO PYRE PHEN BAAN BAPY INPY FLU ANTH DIBA ACEN Alkylphenols NONY

log10 Kp (m3 mg1)

log10 Kd (m3 mg1)

na

Min

P25

P50

P75

Max

na

Min

P25

P50

P75

Max

325 120 250 190 275 265 265 85 95

3.80 3.03 4.03 5.72 4.70 5.75 6.01 4.70 3.12

2.50 2.10 3.24 4.44 3.67 4.70 5.00 4.09 2.42

1.48 1.52 2.79 3.59 3.24 4.34 4.67 3.81 1.84

0.53 0.80 1.79 3.02 2.89 4.04 4.31 3.35 1.27

4.25 1.86 0.13 2.03 1.90 3.27 3.61 1.96 0.35

15 6 15 12 15 12 12 3 6

3.03 1.39 4.54 4.86 4.74 6.06 6.39 3.69 1.89

2.02 0.83 4.22 4.65 4.46 5.91 6.33 3.69 1.75

1.17 0.48 2.48 4.50 4.17 5.57 6.19 3.32 1.19

0.39 0.52 1.89 3.92 3.59 5.13 5.72 3.23 0.98

0.10 0.59 1.40 3.38 3.17 4.79 5.48 3.23 0.92

160 215 150 200 170 120 205 160 110 130

3.41 2.95 2.48 3.04 2.85 2.85 2.39 2.78 1.56 5.41

2.90 2.21 1.52 1.90 1.73 1.64 0.93 1.36 0.39 4.06

2.59 1.75 0.91 1.29 1.16 0.99 0.18 0.57 1.90 3.69

2.24 1.32 0.42 0.63 0.55 0.39 0.66 0.06 3.13 3.24

1.24 0.06 1.03 1.52 1.03 1.03 3.32 1.75 6.70 2.38

18 27 15 30 21 6 24 18 3 9

4.01 3.23 2.16 2.49 2.54 2.18 1.89 1.96 3.75 5.46

3.83 3.06 2.06 2.20 2.08 1.89 1.20 1.39 3.75 5.36

3.62 2.70 1.45 1.74 1.89 1.52 0.90 1.09 5.34 5.23

3.49 2.59 1.37 1.56 1.52 1.09 0.30 0.51 5.42 5.06

3.30 2.31 1.02 1.02 1.02 1.02 0.27 0.27 5.42 4.91

140 230 150 195 120 140 120 185 230 160

5.41 4.58 4.25 4.54 3.70 4.11 3.70 3.90 3.88 3.65

4.04 3.67 3.22 3.26 2.90 3.03 2.89 2.78 2.72 2.31

3.70 3.24 2.93 2.85 2.58 2.62 2.51 2.27 2.19 1.70

3.29 2.92 2.55 2.42 2.17 2.14 2.04 1.79 1.64 1.07

2.38 1.80 1.68 1.20 1.20 1.20 1.20 0.58 0.55 0.12

12 24 15 21 6 12 6 18 24 18

5.38 5.36 4.67 5.00 4.00 4.31 4.00 4.30 4.32 3.59

5.28 4.90 4.13 4.19 3.86 3.99 3.76 3.72 3.69 3.14

5.18 4.67 3.75 3.94 3.64 3.62 3.39 3.24 3.11 2.35

5.09 4.47 3.39 3.75 3.43 3.20 2.71 2.77 2.72 1.98

4.91 4.20 2.96 3.49 3.37 3.00 2.65 2.49 2.22 1.34

155 230 175 175 190 155 155 130 120 120 165 165 185 120 95 140 175 165 120 140 175 175

3.78 3.54 4.21 6.16 6.16 2.97 2.66 2.48 2.79 2.73 4.52 6.40 4.60 7.34 6.05 4.12 3.92 7.96 3.92 4.52 6.28 3.92

2.86 2.76 3.55 4.40 4.61 1.99 1.70 1.53 1.36 2.10 3.55 3.40 3.40 6.21 5.17 3.42 3.47 4.37 3.39 3.50 4.35 3.43

2.36 2.42 3.29 4.05 4.24 1.42 0.84 0.92 0.67 1.71 2.99 2.66 3.06 5.69 4.64 3.08 3.22 3.15 3.02 2.69 4.05 3.18

1.89 1.82 2.89 3.74 3.98 0.77 0.19 0.09 0.07 1.29 2.47 2.04 2.77 5.27 4.12 2.75 2.87 2.40 2.74 2.27 3.72 2.82

0.28 4.23 2.05 1.99 3.09 0.63 1.92 1.92 2.16 0.11 1.52 0.83 2.03 4.63 1.27 2.09 1.75 1.37 2.10 1.52 2.97 1.87

9 24 15 15 12 9 6 6 6 6 12 12 18 6 6 12 15 12 6 12 15 15

2.80 4.17 5.05 6.85 6.85 2.96 3.35 2.74 3.58 3.21 4.75 7.09 5.38 8.03 4.06 4.23 4.52 8.65 3.87 5.10 6.97 4.52

2.31 3.75 4.49 5.37 6.50 2.43 3.03 2.35 3.29 2.97 4.53 6.87 4.72 7.97 3.83 3.99 4.41 7.49 3.83 4.83 5.73 4.28

2.20 3.22 3.96 5.19 5.64 1.79 2.36 2.08 2.53 2.72 4.42 5.62 4.16 7.38 3.38 3.79 4.16 4.83 3.49 3.91 5.45 4.16

1.64 3.02 3.64 4.92 5.38 1.09 1.19 1.36 1.28 2.37 3.57 4.37 3.85 6.92 2.61 3.62 4.07 4.22 3.43 2.98 5.00 3.91

1.55 2.62 3.32 3.26 5.19 0.64 1.12 1.30 1.21 2.31 2.90 4.07 3.57 6.88 2.55 3.36 3.02 3.76 3.37 2.90 4.61 3.14

140 185 150 210 205 245 265 185 220 150 185 185 150 175

2.18 2.41 3.50 3.19 4.08 4.60 5.37 3.33 4.83 1.96 6.21 5.37 2.47 6.35

1.07 1.58 2.12 2.42 3.53 3.52 4.31 2.59 2.25 0.75 4.93 4.38 0.99 5.22

0.34 0.99 1.39 1.52 3.13 2.97 3.96 2.10 1.24 0.13 4.51 3.87 0.30 4.86

0.51 0.20 0.69 0.31 2.68 1.62 3.34 1.55 0.07 1.18 4.11 2.51 1.85 4.45

3.09 3.05 1.09 2.83 1.29 1.65 0.38 0.03 3.16 4.74 3.10 1.39 5.48 3.51

12 18 15 18 24 21 27 18 21 15 18 18 15 15

2.63 2.98 3.19 4.17 4.93 5.38 6.06 4.29 4.29 2.29 6.90 6.06 2.29 7.04

1.99 2.55 2.80 3.81 4.69 5.03 5.79 3.99 3.83 1.79 6.50 5.99 1.69 6.91

1.31 2.33 2.41 3.54 4.52 4.69 5.64 3.83 2.80 1.53 6.36 5.84 1.22 6.78

1.05 2.05 2.29 3.00 4.34 4.40 5.49 3.21 2.23 1.11 6.15 5.52 0.41 6.66

0.45 1.63 1.66 2.56 4.06 4.06 5.32 2.72 1.44 0.57 6.07 5.29 0.67 6.48

120

5.62

4.43

3.88

3.38

2.12

6

4.68

4.52

4.11

3.46

3.40

(continued on next page)

216

W. Wei et al. / Chemosphere 153 (2016) 212e219

Table 3 (continued ) log10 Kp (m3 mg1)

Compound

log10 Kd (m3 mg1) na

Min

P25

P50

P75

Max

Min

P25

P50

P75

Max

120 110

7.44 4.79

5.58 4.07

4.97 3.71

4.48 3.31

3.89 2.49

6 3

5.53 4.15

5.52 4.15

5.45 3.85

5.37 3.77

5.35 3.77

120 130

4.60 5.04

4.13 4.22

3.83 3.76

3.46 3.43

2.61 2.42

6 9

4.60 5.04

4.13 4.22

3.83 3.76

3.46 3.43

2.61 2.42

95 165

5.89 3.33

5.01 2.27

4.57 1.37

4.16 0.58

3.43 1.37

6 12

5.46 1.43

5.40 0.85

5.15 0.47

4.82 0.01

4.76 0.10

n BUTY OCTY Synthetic musks AHTN HHCB Others TBP BPA

a

TRIC: Triclosan; PMET: Permethrin; CYFL: Cyfluthrin; CYPE: Cypermethrin; DMET: Deltamethrin; OXAD: Oxadiazon; DIEL: Dieldrin; a-ENDO: a-endosulfan; CHLO: Chlorpyrifos; DICH: Dichlorvos; METO: Metolachlor; DIAZ: Diazinon; TCHL: trans-chlordane; ALDR: Aldrin; ATRA: Atrazine; ENDR: Endrin; HEPT: Heptachlor; CCHL: cis-Chlordane; BPER: Benzo[g,h,i]perylene; BKFL: Benzo[k]fluoranthene; BBFL: Benzo[b]fluoranthene; CHRY: Chrysene; FLUO: Fluoranthene; PYRE: Pyrene; PHEN: Phenanthrene; BAAN: Benzo[a]anthracene; BAPY: Benzo[a]pyrene; INPY: Indeno[1,2,3-c,d]pyrene; FLU: Fluorene; ANTH: Anthracene; DIBA: Dibenzo[a,h]anthracene; ACEN: Acenaphthene; NONY: 4-n-nonylphenol; BUTY: 4-tert-butylphenol; OCTY: 4-tert-octylphenol; AHTN: Tonalide; HHCB: Galaxolide. a n: number of data.

Kp fom ¼ Kd fom

p d

rd rp

3.3. Distribution of log10 Kp values

(3)

To ensure that Kp remains proportional to Kd, the term fom fom

p d

rd rp

should remain constant. The ratio between Kp and Kd was

obtained by linear regression of the values of log10 Kp and log10 Kd, which were calculated using the retrieved empirical equations. 3. Results and discussion

The calculated values of log10 Kp were considered as real-valued random variables and were fitted to either a normal distribution or a log-normal distribution. The XLSTAT 2014 software (Addinsoft, Paris, France) (see Figs. 1 and 2 as examples) or the Crystal Ball 11.1 software (Oracle, Redwood Shores, CA, USA) were used. The log10 Kp values of 27 SVOCs follow a normal distribution, whereas the log10 Kp values of 45 SVOCs follow a log-normal distribution. The p-value from the Kolmogorov-Smirnov test on the original log10 Kp and

3.1. Empirical equations and available parameters A total of 29 articles associated with the calculation of Kp or Kd were retrieved. Among them, 17 articles gathered 38 original equations to calculate Kp or Kd from p0L or Koa at 25  C. Among the 38 equations, 25 addressed the relationship between log10 Kp and log10 p0L , 10 addressed the relationship between log10 Kp and log10 Koa, and 3 addressed the relationship between log10 Kd and log10 Koa (Table 1). No equations relating Kd and p0L were found. The studies providing log10 p0L or log10 Koa values are displayed in Table 2. On average, 5 log10 p0L values (max ¼ 11 for DEHP, min ¼ 3 for TBP) and 5 log10 Koa values (max ¼ 10 for BDE 99, min ¼ 1 for DMEP) were obtained from the literature and the EPI Suite software for each SVOC (Table S1). 3.2. Percentiles of log10 Kp and log10 Kd values On average, 168 values of log10 Kp (max ¼ 325 for DEHP, min ¼ 85 for DMEP) and 14 values of log10 Kd (max ¼ 30 for BDE 99, min ¼ 3 for DMEP) were calculated for each SVOC. The percentiles of log10 Kp and log10 Kd for each SVOC are reported in Table 3. The 25th percentile (P25), median (P50), and 75th percentile (P75) values of log10 Kp were of the same order of magnitude for 36 SVOCs. The differences between the minimum and maximum values of log10 Kp varied between 1 order of magnitude (atrazine) and 8 orders of magnitude (DEHP). The P25, median, and P75 values of log10 Kd were of the same order of magnitude for 54 SVOCs. The minimum and maximum values of log10 Kd varied between being similar (e.g., DMP) and differing by 4 orders of magnitude (i.e., aldrin). One possible reason for the differences observed between the percentiles is that some retrieved values of p0L and Koa may occasionally deviate from the others. These extreme values were not excluded because (1) less than 11 values of p0L or Koa were available for each SVOC, and (2) there was no scientific criterion to exclude these values.

Fig. 1. Distribution of log10 Kp (m3 mg1) for DBP at 25  C.

Fig. 2. Distribution of log10 Kp (m3 mg1) for BDE 153 at 25  C.

W. Wei et al. / Chemosphere 153 (2016) 212e219 Table 4 Distributions of log10 Kp and parameters of log10 Kd at 25  C. Compound

Table 4 (continued ) Compound

log10 Kp (m3 mg1) AMa

Phthalates DEHP 1.30 DiNP 1.33 BBP 2.50 DiBP 3.74 DBP 3.26 DEP 4.36 DMP 4.67 DMEP 3.68 DOP 1.74 PBDEs 28 2.52 47 1.69 85 0.93 99 1.23 100 1.08 119 1.02 153 0.06 154 0.60 209 1.95 PCBs 28 3.70 31 3.71 52 3.23 77 2.92 101 2.86 105 2.53 118 2.62 126 2.48 138 2.30 153 2.18 180 1.72 Biocides TRIC 2.32 4,40 -DDT 1.94 4,40 -DDE 3.24 g-HCH 4.05 a-HCH 4.29 PMET 1.36 CYFL 0.66 CYPE 0.73 DMET 0.64 OXAD 1.66 DIEL 2.98 a-ENDO 2.87 CHLO 3.08 DICH 5.75 METO 4.37 DIAZ 3.08 TCHL 3.13 ALDR 3.52 ATRA 3.02 ENDR 2.85 HEPT 4.10 CCHL 3.11 PAHs BPER 0.17 BKFL 0.63 BBFL 1.39 CHRY 0.95 FLUO 3.03 PYRE 2.49 PHEN 3.64 BAAN 2.04 BAPY 1.21 INPY 0.38 FLU 4.54 ANTH 3.52 DIBA 0.59 ACEN 4.85 Alkylphenols NONY 3.90

217

log10 Kd (m3 mg1)

SDa

GMb

GSDb

Positionb

DISc

AMa

SDa

1.65 1.05 1.02 0.88 0.57 0.49 0.48 0.61 0.86

3.48 2.32 1.91 e e e e 1.57 2.04

1.52 1.49 1.58 e e e e 1.42 1.46

5.10 3.84 4.61 e e e e 5.36 3.93

L-N L-N L-N N N N N L-N L-N

1.21 0.32 2.96 4.30 4.07 5.52 6.07 3.41 1.32

0.94 0.75 1.16 0.49 0.52 0.43 0.34 0.24 0.39

0.46 0.64 0.74 0.93 0.80 0.87 1.17 1.01 1.98

1.29 2.01 4.81 3.94 6.94 13.11 5.16 14.17 7.84

1.39 1.35 1.16 1.26 1.12 1.07 1.24 1.07 1.27

3.88 3.80 5.79 5.27 8.07 14.16 5.34 14.81 6.12

L-N L-N L-N L-N L-N L-N L-N L-N L-N

3.66 2.78 1.60 1.79 1.84 1.52 0.82 0.95 4.84

0.22 0.27 0.37 0.39 0.45 0.43 0.62 0.70 0.94

0.66 0.64 0.55 0.56 0.69 0.56 0.66 0.58 0.72 0.75 0.88

e e 3.64 e e 5.95 e 11.62 e e e

e e 1.16 e e 1.10 e 1.05 e e e

e e 6.91 e e 8.51 e 14.12 e e e

N N L-N N N L-N N L-N N N N

5.21 5.18 4.71 3.78 4.03 3.65 3.60 3.31 3.26 3.17 2.49

0.18 0.14 0.31 0.50 0.41 0.23 0.42 0.54 0.56 0.61 0.71

0.77 1.44 0.46 0.63 0.52 0.81 1.29 1.07 1.06 0.59 0.72 1.17 0.48 0.62 1.21 0.48 0.45 1.43 0.45 0.77 0.58 0.43

3.66 1.27 4.16 17.74 e 5.71 3.70 e 7.17 2.23 42.52 e 2931.26 e 1.83 e 1.36 e 4.60 e e 1.55

1.23 2.11 1.12 1.04 e 1.15 1.38 e 1.16 1.29 1.02 e 1.00 e 1.70 e 1.36 e 1.10 e e 1.30

6.06 3.61 7.43 21.80 e 7.13 4.55 e 7.88 3.97 45.51 e 2934.34 e 6.48 e 4.56 e 7.65 e e 4.72

L-N L-N L-N L-N N L-N L-N N L-N L-N L-N N L-N N L-N N L-N N L-N N N L-N

2.05 3.35 4.06 5.15 5.83 1.83 2.22 1.96 2.39 2.71 4.11 5.58 4.30 7.43 3.29 3.79 4.05 5.50 3.58 3.93 5.57 4.03

0.42 0.43 0.54 1.10 0.61 0.80 0.95 0.55 1.04 0.33 0.65 1.30 0.58 0.54 0.63 0.25 0.47 1.86 0.21 0.90 0.77 0.41

1.17 1.37 1.03 2.06 0.67 1.45 1.08 0.76 1.71 1.52 0.63 1.12 1.95 0.60

3.70 2.03 e 2.17 1.34 2.12 1.82 3.24 e 3.46 e 2.96 5.17 e

1.34 1.72 e 1.96 1.54 1.73 1.64 1.25 e 1.48 e 1.41 1.41 e

4.14 2.98 e 3.67 4.50 4.95 5.69 5.36 e 3.35 e 6.66 4.90 e

L-N L-N N L-N L-N L-N L-N L-N N L-N N L-N L-N N

1.45 2.32 2.47 3.39 4.51 4.72 5.66 3.64 2.90 1.49 6.37 5.76 1.03 6.77

0.65 0.39 0.42 0.50 0.24 0.39 0.20 0.49 0.85 0.51 0.25 0.26 0.88 0.18

0.76

13.57

1.06

17.49

L-N

4.04

0.54

log10 Kp (m3 mg1) AMa

BUTY 5.07 OCTY 3.70 Synthetic musks AHTN 3.78 HHCB 3.80 Others TBP 4.59 BPA 1.35

log10 Kd (m3 mg1)

SDa

GMb

GSDb

Positionb

DISc

AMa

SDa

0.78 0.52

e 7.33

e 1.07

e 11.05

N L-N

5.45 3.92

0.07 0.20

0.46 0.56

2.14 e

1.23 e

5.96 e

L-N N

4.59 4.69

0.63 0.23

0.56 1.15

e 1.34

e 1.18

e 6.22

N L-N

5.13 0.50

0.29 0.50

TRIC: Triclosan; PMET: Permethrin; CYFL: Cyfluthrin; CYPE: Cypermethrin; DMET: Deltamethrin; OXAD: Oxadiazon; DIEL: Dieldrin; a-ENDO: a-endosulfan; CHLO: Chlorpyrifos; DICH: Dichlorvos; METO: Metolachlor; DIAZ: Diazinon; TCHL: transchlordane; ALDR: Aldrin; ATRA: Atrazine; ENDR: Endrin; HEPT: Heptachlor; CCHL: cis-Chlordane; BPER: Benzo[g,h,i]perylene; BKFL: Benzo[k]fluoranthene; BBFL: Benzo[b]fluoranthene; CHRY: Chrysene; FLUO: Fluoranthene; PYRE: Pyrene; PHEN: Phenanthrene; BAAN: Benzo[a]anthracene; BAPY: Benzo[a]pyrene; INPY: Indeno [1,2,3-c,d]pyrene; FLU: Fluorene; ANTH: Anthracene; DIBA: Dibenzo[a,h]anthracene; ACEN: Acenaphthene; NONY: 4-n-nonylphenol; BUTY: 4-tert-butylphenol; OCTY: 4-tert-octylphenol; AHTN: Tonalide; HHCB: Galaxolide. a AM and SD: Arithmetic mean and standard deviation. b GM and GSD: Geometric mean and standard deviation of shifted distribution, Position: amount of the shift. c DIS: Distribution (N: Normal; L-N: Log-normal).

fitted values is greater than 0.05 for each fit, which indicates that the hypothesis of either normal or log-normal distribution cannot be rejected. For most of the SVOCs, the number of calculated values of log10 Kd is less than 20. Therefore, the distribution of log10 Kd was not fitted in this study. The parameters for each distribution of log10 Kp were fitted using the Crystal Ball software and are shown in Table 4. Notably, other types of distributions (such as Weibull, gamma, and logistic) for some SVOCs may also fit our data with high p-values. These distributions provide similar shapes of curves compared to either normal or log-normal distributions and can be alternatives. However, they were not chosen in this study. The distribution of log10 Kp can be used to predict the distribution of the SVOC concentration in one phase from that in another phase using Eq. (1). Future studies could employ the Monte-Carlo simulation method in the Crystal Ball software, thereby reducing the bias and allowing calculation of the uncertainty in the predictions. 3.4. Span of Kp and Kd values The arithmetic mean and median values of Kp were of the same order of magnitude for the 72 individual SVOCs. The arithmetic mean and median values of Kd were of the same order of magnitude for the 72 SVOCs, except for the two synthetic musks (tonalide and galaxolide). In the case of 60 SVOCs, the arithmetic mean values of Kp were of the same order of magnitude whether Kp was calculated from both or either p0L and Koa (Table S2). The differences in the arithmetic mean values of Kp span over orders of magnitude for 3 phthalates (DiNP, DMEP, and DOP), BDE 209, 3 biocides (triclosan, a-endosulfan, and metolachlor), 4 PAHs (chrysene, pyrene, indeno [1,2,3-c,d]pyrene, and dibenzo[a,h]anthracene), and bisphenol A. This phenomenon can be explained with two possible reasons. First, some of the retrieved p0L and Koa values for a given compound occasionally deviate substantially from the other values. For example, the retrieved p0L values for di-isononyl phthalate (DiNP) and bisphenol A span 2 orders of magnitude. Because fewer than 11 values of p0L and Koa are available for each SVOC, no scientific technique can be used to verify or discard any value. All retrieved values of p0L and Koa were taken into account in the calculation.

218

W. Wei et al. / Chemosphere 153 (2016) 212e219

Program on Endocrine Disruptors (PNRPE; Grant n 2100522667), the French Agency for Food, Environmental and Occupational Health and Safety (ANSES; Grant n 2011-1-128), the French Observatory of Indoor Air Quality (OQAI; Grants 2011 and 2012), the Scientific and Technical Center for Building (CSTB), and the School of Public Health (EHESP). The preparation of this manuscript was performed during a scientific visit to CSTB (WW) and was supported by a grant from the “Carnot Programme” (Grant 2011). Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.chemosphere.2016.03.007. References

Fig. 3. Linear regression of the median values of log10 Kp and log10 Kd (m mg 25  C. 3

1

) at

Notably, scattered values of p0L and Koa may have affected the determined distribution. Therefore, in future studies, the accuracy of the p0L and Koa values will need to be assessed and methods to detect and discard outliers will need to be developed. The second possible reason is that some empirical equations were originally developed for a specific chemical family of SVOCs and were assumed to be applicable to all target SVOCs. This assumption may result in larger errors in the calculation of Kp or Kd for select SVOCs. A systematic study of Kp and Kd for all SVOC families may substantially reduce this bias. 3.5. Relationship between Kp and Kd The linear regression showed that the median values of log10 Kp and log10 Kd were proportional for the 72 SVOCs (Fig. 3). Therefore, an empirical relationship between Kp and Kd is estimated as follows:

Kp ¼ 8:32 Kd

(4)

The ratio of Kp and Kd for the 72 SVOCs is 8.32. Comparing Eq. (3) and Eq. (4), Weschler and Nazaroff (2010) proposed that the ratio of Kp and Kd, as calculated using the values of fom p , fom d , rp , and rd , is 4. The values of fom p , fom d , rp , and rd were not simultaneously measured in the same environment. Therefore, their values may have been influenced by environmental factors, such as smoking and non-smoking (Fromme et al., 2005). The error in calculating the ratio of Kp and Kd using Eq. (4) remains unknown. 4. Conclusions SVOC particle/gas and dust/gas partition coefficients are useful for predicting SVOC concentrations in the gas-phase from the SVOC concentrations in airborne particles or settled dust. Published empirical equations and parameters enabled us to determine the distributions of log10 Kp and log10 Kd values for 72 SVOCs. These distributions may serve as references with Monte-Carlo simulations to predict SVOC concentrations in one phase from those in another phase. Acknowledgments The ECOS project was supported by the French Scientific

Baker, J.E., Eisenreich, S.J., 1990. Concentrations and fluxes of polycyclic aromatic hydrocarbons and polychlorinated biphenyls across the air-water interface of Lake Superior. Environ. Sci. Technol. 24, 342e352. Benning, J.L., Liu, Z., Tiwari, A., Little, J.C., Marr, L.C., 2013. Characterizing gas-particle interactions of phthalate plasticizer emitted from vinyl flooring. Environ. Sci. Technol. 47, 2696e2703. Chen, J.W., Harner, T., Yang, P., Quan, X., Chen, S., Schramm, K.W., Kettrup, A., 2003. Quantitative predictive models for octanol-air partition coefficients of polybrominated diphenyl ethers at different temperatures. Chemosphere 51, 577e584. Cotham, W.E., Bidleman, T.F., 1995. Polycyclic aromatic hydrocarbons and polychlorinated biphenyls in air at an urban and rural site near Lake Michigan. Environ. Sci. Technol. 29, 2782e2789. Finizio, A., Mackay, D., Bidleman, T., Harner, T., 1997. Octanol-air partition coefficient as a predictor of partitioning of semi-volatile organic chemicals to aerosols. Atmos. Environ. 31, 2289e2296. Foreman, W.T., Bidleman, T.F., 1990. Semivolatile organic compounds in the ambient air of Denver, Colorado. Atmos. Environ. 24, 2405e2416. Fromme, H., Lahrz, T., Hainsch, A., Oddoy, A., Piloty, M., Rüden, H., 2005. Elemental carbon and respirable particulate matter in the indoor air of apartments and nursery schools and ambient air in Berlin (Germany). Indoor Air 15, 335e341. Harner, T., 1998. Octanol-air partition coefficient for describing particle/gas partitioning of aromatic compounds in urban air. Environ. Sci. Technol. 32, 1494e1502. Harner, T., Mackay, D., 1995. Measurement of octanol-air partition coefficients for chlorobenzenes, PCBs, and DDT. Environ. Sci. Technol. 29, 1599e1606. He, J., Balasubramanian, R., 2009. A study of gas/particle partitioning of SVOCs in the tropical atmosphere of Southeast Asia. Atmos. Environ. 43, 4375e4383. Hoff, R.M., Strachan, W.M.J., Sweet, C.W., Chan, C.H., Shackleton, M., Bidleman, T.F., Brice, K.A., Burniston, D.A., Cussion, S., Gatz, D.F., Harlin, K., Schroeder, W.H., 1996. Atmospheric deposition of toxic chemicals to the Great Lakes: a review of data through 1994. Atmos. Environ. 30, 3505e3527. Kaupp, H., McLachlan, M.S., 1999. Gas/particle partitioning of PCDD/Fs, PCBs, PCNs and PAHs. Chemosphere 38, 3411e3421. Kaupp, H., Umlauf, G., 1992. Atmospheric gas-particle partitioning of organic compounds: a comparison of sampling methods. Atmos. Environ. 26, 2259e2261. Li, X.H., Chen, J.W., Zhang, L., Qiao, X.L., Huang, L.P., 2006. The fragment constant method for predicting octanol-air partition coefficients of persistent organic pollutants at different temperatures. J. Phys. Chem. Ref. Data 35, 1365e1384. Ligocki, M.P., Pankow, J.F., 1989. Measurements of the gas/particle distribution of atmospheric organic compounds. Environ. Sci. Technol. 23, 75e83. Mandin, C., Mercier, F., Lucas, J.P., Ramalho, O., Blanchard, O., Bonvallot, N., Raffy, G., Gilles, E., Glorennec, P., Le Bot, B., 2014a. ECOS-POUSS: a nationwide survey of semi-volatile organic compounds in home settled dust. In: Proceeding of Indoor Air Conference, Hongkong, HP1015. Mandin, C., Mercier, F., Lucas, J.P., Ramalho, O., Gilles, E., Blanchard, O., Bonvallot, N., Glorennec, P., Le Bot, B., 2014b. ECOS-PM: a nationwide survey of semi-volatile organic compounds in indoor air. In: Proceeding of Indoor Air Conference, Hongkong, HP1017. Naumova, Y.Y., Offenberg, J.H., Eisenreich, S.J., Meng, Q.Y., Polidori, A., Turpin, B.J., Weisel, C.P., Morandi, M.T., Colome, S.D., Stock, T.H., Winer, A.M., Alimokhtari, S., Kwon, J., Maberti, S., Shendell, D., Jones, J., Farrar, C., 2003. Gas/particle distribution of polycyclic aromatic hydrocarbons in coupled outdoor/indoor atmospheres. Atmos. Environ. 37, 703e719. Ngabe, B., Bidleman, T.F., 1992. Occurrence and vapour particle partitioning of heavy organic compounds in ambient air in Brazzaville, Congo. Environ. Pollut. 76, 147e156. Odabasi, M., Cetin, B., 2012. Determination of octanol-air partition coefficients of organochlorine biocides (OCPs) as a function of temperature: application to airsoil exchange. J. Environ. Manage. 113, 432e439. Odabasi, M., Cetin, B., Sofuoglu, A., 2006. Determination of octanol-air partition coefficients and supercooled liquid vapor pressures of PAHs as a function of

W. Wei et al. / Chemosphere 153 (2016) 212e219 temperature: application to gas-particle partitioning in an urban atmosphere. Atmos. Environ. 40, 6615e6625. Pankow, J.F., 1994. An absorption model of gas/particle partitioning of organic compounds in the atmosphere. Atmos. Environ. 28, 185e188. Salthammer, T., Schripp, T., 2015. Application of the Junge- and Pankow-equation for estimating indoor gas/particle distribution and exposure to SVOCs. Atmos. Environ. 106, 467e476. Shoeib, M., Harner, T., 2002. Using measured octanol-air partition coefficients to explain environmental partitioning of organochlorine biocides. Environ. Toxicol. Chem. 21, 984e990. Shoeib, M., Harner, T., Wilford, B.H., Jones, K.C., Zhu, J.P., 2005. Perfluorinated sulfonamides in indoor and outdoor air and indoor dust: occurrence, partitioning, and human exposure. Environ. Sci. Technol. 39, 6599e6606. United States Environmental Protection Agency Website. http://www2.epa.gov/ tsca-screening-tools/download-epi-suitetm-estimation-program-interface-

219

v411. Last access on Dec. 2nd, 2015. Weschler, C.J., 2003. Indoor/outdoor connections exemplified by processes that depend on an organic compound's saturation vapor pressure. Atmos. Environ. 37, 5455e5465. Weschler, C.J., Nazaroff, W.W., 2008. Semivolatile organic compounds in indoor environments. Atmos. Environ. 42, 9018e9040. Weschler, C.J., Nazaroff, W.W., 2010. SVOC partitioning between the gas phase and settled dust indoors. Atmos. Environ. 44, 3609e3620. Weschler, C.J., Salthammer, T., Fromme, H., 2008. Partitioning of phthalates among the gas phase, airborne particles and settled dust in indoor environments. Atmos. Environ. 42, 1449e1460. Yamasaki, H., Kuwata, K., Miyamoto, H., 1982. Effect of ambient temperature on aspects of airborne polycyclic aromatic hydrocarbons. Environ. Sci. Technol. 16, 189e194.