Aqueous solubility, HenryХs law constants and air
Transcription
Aqueous solubility, HenryХs law constants and air
Chemosphere 57 (2004) 1543–1551 www.elsevier.com/locate/chemosphere Aqueous solubility, HenryÕs law constants and air/water partition coefficients of n-octane and two halogenated octanes S. Sarraute, H. Delepine, M.F. Costa Gomes *, V. Majer Laboratoire de Thermodynamique des Solutions et des Polyme`res, UMR 6003, CNRS/Univ. Blaise Pascal, 24, avenue des Landais, 63177 Aubie`re Cedex, France Received 16 July 2003; received in revised form 19 July 2004; accepted 22 July 2004 Abstract New data on the aqueous solubility of n-octane, 1-chlorooctane and 1-bromooctane are reported between 1 C and 45 C. HenryÕs law constants, KH, and air/water partition coefficients, KAW, were calculated by associating the measured solubility values to vapor pressures taken from literature. The mole fraction aqueous solubility varies between (1.13– 1.60) · 107 for n-octane with a minimum at approximately 23 C, (3.99–5.07) · 107 for 1-chlorooctane increasing monotonically with temperature and (1.60–3.44) · 107 for 1-bromooctane with a minimum near 18 C. The calculated air–water partition coefficients increase with temperature and are two orders of magnitude lower for the halogenated derivatives compared to octane. The precision of the results, taken as the average absolute deviations of the aqueous solubility, the HenryÕs law constants, or the air/water partition coefficients, from appropriate smoothing equations as a function of temperature is of 3% for n-octane and of 2% and 4% for 1-chlorooctane and 1-bromooctane, respectively. A new apparatus based on the dynamic saturation column method was used for the solubility measurements. Test measurements with n-octane indicated the capability of measuring solubilities between 106 and 1010 in mole fraction, with an estimated accuracy better than ±10%. A thorough thermodynamic analysis of converting measured data to air/water partition coefficients is presented. 2004 Elsevier Ltd. All rights reserved. Keywords: Aqueous solubility; HenryÕs law constant; Air/water partition coefficient; n-Octane; 1-Chlorooctane; 1-Bromooctane 1. Introduction The prediction of the transport and fate of hydrophobic organic chemicals in the environment requires * Corresponding author. Fax: +33 4 73407185. E-mail address: margarida.c.gomes@univ-bpclermont.fr (M.F. Costa Gomes). the knowledge of their physical and chemical properties. An important thermodynamic property is the solubility in water as it is essential in waste minimization and waste remediation calculations. Furthermore, when combined with vapor pressure data, it permits the calculation of the HenryÕs law constant and from there, several partition coefficients such as the air/water partition coefficient or the octanol/water coefficient (Mackay, 1991; Boethling and Mackay, 2000). 0045-6535/$ - see front matter 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2004.07.046 1544 S. Sarraute et al. / Chemosphere 57 (2004) 1543–1551 The solubility of a pollutant in water constitutes per se a physical property of capital importance for environmental processes. It can affect adsorption and desorption in soils and volatility of chemicals from aquatic systems and it has also impact in possible transformations by hydrolysis, photolysis, oxidation, reduction and biodegradation. Other specialized pathways, such as washout from atmosphere by rain, are also influenced by the extent of solubility (Reza et al., 1999). The air–water partition coefficient, KAW, relating atmospheric and aqueous concentrations of a substance is in direct relation with HenryÕs law constant, KH, which can be calculated using solubility data. Apart from their need in the construction of transport models that attempt to describe the movement of pollutants in the environment, these data are also required in a number of technological applications (e.g. air-stripping processes for the renovation of waters contaminated by organic solvents (Gosset, 1987)). Different methods, extensively reviewed by Mackay and Shiu (1981) were described for the determination of HenryÕs law constants. More recently, several other techniques have been proposed and their respective merits and deficiencies discussed (Sanemasa et al., 1982; Gosset, 1987; Dohnal and Hovorka, 1999). Halogenated paraffins, in particular chlorinated alkanes are used as plasticizers, lubricants and flame retardants (Sijm and Sinnige, 1995). Very little information is available on the physical–chemical properties of these persistent hydrophobic chemicals such as aqueous solubility, octanol/water partition coefficient, HenryÕs law constant, relative bioavailability, while more information is available on their acute toxicity and bioaccumulation (Drouillard et al., 1998a). The complexity of technical mixtures of chlorinated paraffins has precluded the determination of physical properties of individual compounds or of simpler synthesized mixtures to determine key physical properties that may be used in assessing the behavior of halogenated paraffins in the environment (Drouillard et al., 1998a), namely as a prerequisite for many thermodynamics-based environmental fate models in use today (Drouillard et al., 1998b). In this work, the solubilities of n-octane, 1-chlorooctane and 1-bromooctane in water were determined experimentally as a function of temperature using a dynamic saturation column method. The data are associated with the pure solute vapor pressures for the calculation of the HenryÕs law constant and air/water partition coefficient. The thermodynamic background and simplifying assumptions used in the conversion of the experimental solubility data and in the relevant coefficients are discussed and possible sources of error are outlined. Several data have been reported in the literature concerning the aqueous solubility of n-octane (Polak and Lu, 1953; McAuliffe, 1963, 1966; Price, 1976; Jo¨nsson et al., 1982; Tewari et al., 1982; Heidman et al., 1985; Reza et al., 1999; Miller and Hawthorne, 2000) so this system was chosen to test the experimental technique and also as a reference for examining the effect of the halogen atoms on the aliphatic chain. For 1-chlorooctane no experimental values for the aqueous solubility were found in the literature. For 1-bromooctane two identical experimental values for the aqueous solubility at 25 C, determined using a generator column procedure, are reported independently by Tewari et al. (1982) and Miller and Wasik (1985). These results are presented as a reliable recommended value (xs = 1.56 · 107) in the compilation by Horvath and Getzen (1999) and are also referred in the review paper of Abraham (1984) on solution thermodynamics of organic compounds. Values for the air/water partition coefficients (the so called dimensionless HenryÕs law constants) of the three studied solutes are also reported by Yaffe et al. (2003) for testing several prediction methods. We could not, however, trace the direct experimental determinations of the aqueous solubility or other properties that served as a basis for the calculation of these data. 2. Experimental section 2.1. Materials All the solutes used were obtained from Acros Organics with stated purities of at least 99 mol% and were used with no further purification. Distilled water was used as solvent. 2.2. Apparatus and operation The experimental technique used in this work is based on a dynamic saturation column method and, in its essence, was already described in the literature for use at ambient (May and Wasik, 1978; DeVoe et al., 1981; Owens et al., 1986) and superambient conditions (Miller and Hawthorne, 1998; Bergin, 2002). It is based on the saturation of a known quantity of water flowing through a saturation cell which is filled with an inert stationary phase impregnated with the organic solute. The solute is subsequently trapped in a specific extraction column and finally quantitatively removed by an appropriate solvent. This method was originally proposed for solids and later adapted for liquid solutes (Owens et al., 1986). The analysis of the final solution are carried out either using liquid (May and Wasik, 1978; DeVoe et al., 1981; Owens et al., 1986) or gas chromatography (Miller and Hawthorne, 1998; Bergin, 2002). The apparatus used is schematically represented in Fig. 1. The heated components are kept inside a water bath maintained at constant temperature to within S. Sarraute et al. / Chemosphere 57 (2004) 1543–1551 1545 Fig. 1. Solubility apparatus used in this work. VP, isocratic pump; M, manometer; SC, saturation cell preceded by a pre-heating coil; EC, extraction cell; T, liquid thermostat. ±0.05 C by means of a PID temperature controller (TRONAC Inc., model PTC-40). A pump operating in a constant-flow mode (Gilson, model 306 or Perkin Elmer, model 410 were employed successively) was used during the measurements to supply a regular stream of water required for the solubility determinations. The solvent is pumped during 10-12 h, at flow rates between 0.1 and 0.2 ml/min, through a pre-heating coil (of approximately 2 ml), into the saturation cell (a 2.0 ml, 0.46 cm i.d. · 10 cm analytical column from Modulo-cart) containing the solute immobilized in a dry and clean solid stationary phase (Gaz chrom R 60/80 by Alltech). The saturated solution then flows through an inverse phase analytical column (C18 5lm, 0.46 cm i.d. · 10 cm, from Kromacil) where the solute is quantitatively retained. The pressure, read at the pump, varies from 40 to 70 bar (and decreases 20–30 bar along the saturation cell), depending on the temperature and on the degree of the column packing (different fillings of the saturation cell corresponding to different pressure drops in relation with the packing of the solid sorbent). The organic compound is subsequently removed by back flushing this extraction column with a good solvent (5-15 ml of methanol in our case). The quantity of water passed through the saturation (and extraction) cell, typically 100–200 ml, is determined gravimetrically. The analyses of the final solutions were performed using a Shimadzu model CG14A gas chromatograph equipped with FID and a split/splitless injector. Chromatographic separations were accomplished with a 30 m capillary column (from SGE, 0.32 mm i.d., 0.25 lm film thickness). The quantification of the studied organic compounds was carried out by an internal standard technique using n-nonane, n-decane and n-dodecane for the analysis of n-octane and of the chloro and bromo octane, respectively. To determine whether saturation of the water was achieved and to verify that the solute was not mechanically carried away out of the saturation cell, determinations were performed at different flow rates (0.1–0.3 ml/ min) for all the solutes studied. Identical results were obtained for each flow rate, demonstrating that the residence time of the solvent in the saturation cell was sufficient to saturate the water with the organic solutes. Two extraction columns of different lengths (5 and 10 cm) were used in order to test if the solutes are quantitatively captured. The quantity of methanol used during the back flush of the column was also optimised to assure the complete extraction of the organic solute. 2.3. Thermodynamic data analysis The solubility of the organic solutes in water, expressed as molar fraction of the solute or in quantity of solute per volume of solvent, is calculated from the experimental quantities: the amount of water passed through the saturation column and the amount of solute retained in the specific extraction column, determined by gas chromatography. As mentioned above, the resistance offered by the packing of the saturation cell implies that the saturation of the water stream by a solute is done at an elevated pressure, which decreases along the saturation cell. Since the solubility data of interest to environmental studies is close to ambient pressure (p0 = 1 bar) it will be examined to what extent the measured values of aqueous solubility and HenryÕs law constant are affected by the pressure attained during the experiments. To our knowledge, this issue has not been addressed by earlier investigators who have also used the saturation column method. A detailed discussion of the thermodynamic relations used in the data analysis presented below can be found in reference text books of chemical thermodynamics (Smith et al., 1996; Prausnitz et al., 1999; Sandler, 1999). Specifically, the thermodynamic treatment of the HenryÕs law constant in relation with the aqueous solubility of organic compounds was recently addressed by Sedlbauer et al. (2002). The dependence of the mole fraction solubility xsol s with pressure can be approximated by a differential equation: 1546 S. Sarraute et al. / Chemosphere 57 (2004) 1543–1551 o ln xsol V 1 V s s ffi s op RT T ð1Þ where V 1 s and V s are the partial molar volume of the solute in water at infinite dilution and the molar volume of the pure solute, respectively. This equation is exact in the limit of infinite dilution where the HenryÕs activity coefficient equals unity and no water is dissolved in the organic phase. In the temperature range of the present experiments, both V 1 s and V s do not vary significantly with pressure and the relative error dr xsol (expressed in s percentage), due to neglecting the pressure dependence of the mole fraction solubility, can be estimated as: dr xsol s ¼ dxsol s 100 sol xs ðp0 Þ ðV 1 V s Þðp p0 Þ 1 100 ffi exp s RT ð2Þ 0 0 where xsol s ðp Þ is the mole fraction solubility at p = 1 bar. The difference between the partial molar volume at infinite dilution and the molar volume of a pure liquid hydrophobic solute is generally negative near 25 C. We suppose, by analogy with accurate data available for benzene and toluene (Degranges, 1998) (for which 1 3 V1 s V s ¼ 10 cm mol ), that for the solutes treated, 1 V s does not exceed –15% of V s and that the pressure attained at the end of the column, will never surpass 40 bar. The systematic error in the solubility determination, due to the pressure during the measurements, will then be positive and less than 3% of the mole fraction. The exact thermodynamic definition of the HenryÕs law constant, KH, is: fs ðp; T ; xs Þ ð3Þ K H ðp; T Þ ¼ lim xs !0 xs where fs is the fugacity of the organic solute in the aqueous phase and xs is the corresponding mole fraction. HenryÕs law constant has always dimension of pressure; it should not be considered as a coefficient characterizing phase equilibrium but rather as a proportionality constant between the fugacity of a solute and its concentration in the same phase in the limit of infinite dilution. For hydrophobic substances, sparingly soluble in water, the limit in Eq. (3) can be approximated by the ratio fs =xsol s . In this case, the fugacity of the solute in the aqueous phase is equal to that in the organic phase as we are in a situation of thermodynamic equilibrium. For the organic liquids investigated in this study, the fugacity of the solute in the organic phase can be approximated by the fugacity of the pure solute, fs , which is then calculated in the usual way V s ðp psat sat sat sol s Þ fs ðp; T ; xs Þ ffi fs ðp; T Þ ffi /s ps exp ð4Þ RT where psat s and V s are the vapor pressure and the molar volume of the pure solute, respectively. The exponential term in Eq. (4) is the so called Poynting correction which expresses the change in fugacity due to a variation of pressure between psat s and p (e.g. the pressure of the solubility measurement), provided that V s can be considered constant. The fugacity coefficient of the solute is obtained from (at temperatures below or around the normal boiling point of the solute): sat Bs ps ð5Þ /sat ¼ exp s RT where Bs is the second virial coefficient for the pure solute. The HenryÕs law constant at the experimental pressure can then be calculated from the measured mole fraction solubilities, xsol s , and the thermodynamic properties of the pure solute: V s ðp psat s Þ sat /sat s p s exp RT ð6Þ K H ðp; T Þ ffi xsol s HenryÕs law constant is in direct relation with the 1 ig 0 Gibbs free energy of hydration DG1 hyd ¼ Gs Gs ðp Þ ¼ 0 RT lnðK H =p Þ which corresponds to the difference in the Gibbs free energy of the solute in the standard state of infinite dilution and as an ideal gas at the pressure p0. The pressure dependence of KH is then expressed by o ln K H V1 ð7Þ ¼ s op RT T Assuming that the partial molar volume at infinite dilution of the solute, V 1 s , is independent of pressure, the HenryÕs law constant at the reference pressure p0 can be calculated by V s p psat s sat /sat p exp s s RT 1 ð8Þ K H ðp0 ; T Þ ffi V ð p p0 Þ s xsol exp s RT For the temperatures studied, the vapor pressure of the solutes is low (<0.1 bar), the fugacity coefficient is close to unity and the HenryÕs law constant can be approximated as K H ðp0 ; T Þ ffi psat s xsol s ð9Þ the comparison of Eqs. (2) and (8) Since p > p0 psat s suggests that neglecting both exponential terms in Eq. (8) introduces an uncertainty similar to that estimated above for the solubility. This systematic error is, however, of opposite sign, i.e. the HenryÕs law constant calculated using Eq. (9) will be lower by less than 3% due to the relatively high pressure in the saturation cell. HenryÕs law constant is often used to express the air– water partition coefficient which is defined as (Mackay, 1991) S. Sarraute et al. / Chemosphere 57 (2004) 1543–1551 K AW ¼ lim aq C s !0 C air s C aq s ð10Þ i.e. the limiting ratio of the concentration of a solute, expressed in molarity, in the atmospheric and aqueous phases. In the majority of the conditions relevant to environmental studies, the two coexisting phases are in equilibrium, the organic solute is very dilute in the aqueous phase and air can be considered as an ideal gas (the fugacity of the solute is then equal to its partial pressure ps). In these conditions, sol C air s ffi p s =RT ffi K H xs =RT C aq s ffi xsol s V w ð11Þ The limit in Eq. (10) can then be replaced by the ratio of concentrations expressed in molarity leading to the correct relation between the air/water partition coefficient and HenryÕs law constant: K AW ¼ K HM w RT qw ð12Þ where KH is the HenryÕs law constant expressed in Pa, R is the gas constant expressed in J mol1 K1, Mw and qw are the molar mass of water in kg mol1 and its density in kg m3, respectively. The uncertainty in the calculated KAW will be similar to that estimated for KH as the properties of pure water are very well known. Considering the thermodynamic treatment presented and the significance of the HenryÕs law constant which cannot be considered as a partition coefficient, the authors recommend that the air/water partition coefficient should not be associated with the term HenryÕs law constant. 3. Results and discussion The present experimental method was tested by measuring the solubility of n-octane in water from 10 to 45 C. The experimental results, expressed in molarity and mole fraction, are recorded in Table 1 together with the calculated HenryÕs law constants and air–water partition coefficients. For the vapor pressure of n-octane was used the recommended value of Ruzicka and Majer (1994), the relative atomic masses were taken from the IUPAC tables (IUPAC, 1995) and the orthobaric properties of water from the Wagner and Pruss equation of state (Wagner and Pruss, 1993). The mole fraction aqueous solubilities obtained in this work for n-octane are represented in Fig. 2 together with literature data reported in the same temperature range. The values published by different authors exhibit a large scatter, typical of the difficulties encountered in the measurement of such small values of solubility. A 1547 solubility minimum is observed as expected for an alkane of medium molar mass. The exact value of the temperature corresponding to that minimum is, however, difficult to locate due to the scatter of the experimental values. From the analysis of all these data as well as a careful study of the sources of the systematic errors during our experiments, it is believed that the values for the mole fraction aqueous solubility reported in this work are accurate to within 10%. The results obtained for the aqueous solubility of 1-chlorooctane and 1-bromooctane between 5 and 40 C and 1 and 40 C, respectively, are depicted in Fig. 3 and recorded in Table 1 together with the calculated HenryÕs law coefficients and air–water partition coefficients. The vapor pressures for the halogenated alkanes necessary for the calculation of the HenryÕs law constants, were obtained from the Antoine equation (Smith et al., 1996) with parameters taken from the work of Li and Rossini (1961) (which reports Antoine constants above 65 C for 1-chlorooctane and above 70 C for 1-bromooctane). The validity of the extrapolation of the Antoine equation towards lower temperatures was verified by comparing the enthalpies at 25 C calculated from the Clapeyron equation with the ones determined from calorimetric experiments (Majer and Svoboda, 1985). It was observed that the derived enthalpies of vaporization did not differ more than 1.7% from the experimental values which is considered as satisfactory. The temperature dependence of the aqueous solubility of the three organic solutes was fitted to polynomial equations of the general form: ln xsol s ¼ n X Ai ðT Þi ð13Þ i¼0 Several empirical methods were reported in the literature to represent the temperature dependence of the HenryÕs law constant. In the present case, KH is fitted to a linear function of the inverse temperature (Krause and Benson, 1989): ln K H ¼ n X Bi ð1=T Þi ð14Þ i¼0 and equivalently for the air/water partition coefficient: ln K AW ¼ n X C i ð1=T Þi ð15Þ i¼0 The coefficients Ai, Bi and Ci for Eqs. (13)–(15) as well as the average absolute per-cent deviations of the fit, are listed in Table 2. Three parameter equations were used as is current practice in the literature. It is observed that the errors associated with the parameters of the fit never exceed the absolute value of the parameter. Furthermore, the correlation factors are always acceptable. The use of more than three parameters in such limited 1548 S. Sarraute et al. / Chemosphere 57 (2004) 1543–1551 Table 1 Aqueous solubility (expressed in molarity and mole fraction), vapor pressure, HenryÕs law constant and air–water partition coefficient of n-octane, chlorooctane and bromooctane T/C n-Octane 10.3 24.8 24.9 24.9 24.9 24.9 24.9 24.9 30.1 34.9 35.0 39.9 39.9 44.7 44.8 44.8 Cs/106 mol dm3 7.55 6.87 6.84 7.31 7.14 6.99 7.20 7.49 6.24 8.13 7.34 8.04 8.20 8.67 8.64 8.79 7 xsol s =10 psat/Paa KH/108 Pa KAW 1.36 1.24 1.24 1.32 1.29 1.26 1.30 1.35 1.13 1.47 1.33 1.46 1.49 1.58 1.57 1.60 762.0 1841 1846 1846 1846 1846 1846 1846 2472 3185 3210 4114 4114 5224 5237 5249 56.0 148 149 140 143 146 142 136 219 216 241 282 276 331 333 328 42.8 108 109 102 104 107 103 99.5 157 153 171 197 193 228 229 226 1-Chlorooctane 5.0 9.9 9.9 9.9 10.0 19.1 25.0 25.1 25.2 30.0 34.8 35.1 35.1 40.0 22.7 23.7 23.5 22.6 22.1 25.4 24.7 26.2 25.8 27.1 27.0 26.9 28.0 27.5 4.09 4.27 4.23 4.08 3.99 4.59 4.46 4.73 4.67 4.90 4.90 4.87 5.07 4.99 26.30 39.55 39.71 39.71 40.04 81.99 126.6 127.1 128.0 179.8 248.4 252.5 252.5 347.5 0.51 0.75 0.76 0.79 0.82 1.51 2.46 2.33 2.37 3.23 4.53 4.63 4.45 6.29 0.40 0.58 0.58 0.61 0.62 1.12 1.79 1.70 1.73 2.32 3.20 3.28 3.15 4.39 1-Bromooctane 1.1 5.0 5.2 9.9 9.9 14.9 24.9 29.9 34.9 40.0 40.1 13.8 11.5 12.5 9.11 9.18 9.41 8.88 11.3 14.4 18.9 18.0 2.48 2.07 2.26 1.64 1.66 1.70 1.60 2.05 2.60 3.44 3.27 6.560 9.431 9.606 14.68 14.68 22.60 50.53 73.61 105.5 150.1 150.6 0.27 0.45 0.43 0.89 0.89 1.33 3.15 3.59 4.06 4.37 4.60 0.21 0.35 0.33 0.68 0.68 1.00 2.30 2.58 2.87 3.05 3.21 a From Ruzicka and Majer (1994) for n-octane and from Li and Rossini (1961) for 1-chlorooctane and 1-bromooctane. The values in italics are extrapolations using Antoine equation as explained in the text. temperature range leads always to overfitting without decreasing significantly the average absolute deviation of the fit. Consequently, very large errors in the parameters of Eqs. (13)–(15) were observed. The average absolute deviations listed characterize the precision of the results obtained and vary from 2% to 4% for the three solutes studied. By taking the temperature derivative of Eq. (9), one obtains: R d ln K H d ln xs dp ¼ RT 2 RT 2 sat dð1=T Þ dT dT ð16Þ When the vapor phase is considered as ideal, this equation is approximately equivalent to S. Sarraute et al. / Chemosphere 57 (2004) 1543–1551 2.0 1549 6 1.5 x /10-7 xoctane /10-7 5 4 3 1.0 2 0.5 10 20 30 40 50 1 0 T /°C 20 30 40 50 T /°C Fig. 2. Mole fraction solubility of n-octane in water. (d) Mole fraction solubility, this work; other symbols stand for mole fraction solubilities from literature: (h) McAuliffe (1966); (,) Polak and Lu (1953); () Price (1976); ( ) Tewari et al. (1982); (n) Jo¨nsson et al. (1982); (m) Heidman et al. (1985); (.) Reza et al. (1999); () Miller and Hawthorne (2000). Dhyd H 1 ¼ Dsol H 1 Dvap H ig 10 ð17Þ where the term on the left hand side is the enthalpy of hydration which corresponds to the transfer of the solute from an ideal gas phase to the infinitely dilute solution at the same temperature. The term DsolH1 concerns the transfer of the pure solute to the infinitely dilute solution and DvapHig is the enthalpy of vaporization of the pure solute to an ideal gas state. By taking the temperature derivatives of Eqs. (13) and (14), the values ob- Fig. 3. Mole fraction aqueous solubilities. (j) Chlorooctane, this work; (d) bromooctane, this work; (s) bromooctane, Horvath and Getzen (1999). tained for the solubility and the HenryÕs law constant can be confronted by means of Eq. (17) with the tabulated calorimetric values of DvapHig at 25 C (Majer and Svoboda, 1985). This comparison is possible because, as shown above, the enthalpy of vaporization calculated from vapor pressures using the Clapeyron equation is close to the calorimetric value. Thus, by using Eq. (17) the validity of the temperature correlation of the present values of KH and xs are tested. This is shown in Table 3 where the last column represents the percentage deviation of DvapHig calculated from Eq. (17) from the calorimetric values. It is apparent that in Table 2 Coefficients Ai for Eq. (13) Bi for Eq. (14) and Ci for Eq. (15), for mole fraction solubility, HenryÕs law constants and air/water partition coefficients, respectively, and average per cent absolute deviations (AADa) from fits n-Octane A0 A1 A2 AAD +25.526 2.7975 · 101 +4.7251 · 104 3.1% B0 B1 B2 AAD 7.0475 +2.2996 · 104 4.1490 · 106 3.3% C0 C1 C2 AAD 29.139 +2.4687 · 104 4.3548 · 106 3.1% 1-Chlorooctane A0 A1 A2 AAD 21.845 +4.2442 · 102 6.0619 · 105 2.1% B0 B1 B2 AAD +33.573 2.3949 · 103 5.5673 · 105 2.1% C0 C1 C2 AAD +13.927 2.1286 · 103 5.5538 · 105 2.1% 1-Bromooctane A0 A1 A2 AAD +122.13 9.4611 · 101 1.6240 · 103 4.4% B0 B1 B2 AAD 110.56 +8.2195 · 104 1.2945107 4.4% C0 C1 C2 AAD 130.20 +8.2456 · 104 1.2943 · 107 4.3% a Calculated as P AAD ¼ jðX exp X calc =X exp j n where Xexp is the experimental value, Xcalc is the calculated number and n is the number of points fitted. 1550 S. Sarraute et al. / Chemosphere 57 (2004) 1543–1551 Table 3 Enthalpy of hydration, DhydH1, enthalpy of solution, DsolH1 and ideal enthalpy of vaporization, DvapHig of the three solutes Substance DhydH1/kJ mol1 DsolH1/kJ mol1 DvapHig/kJ mol1a Deviation/% n-Octane 1-Chlorooctane 1-Bromooctane 40.2 51.0 38.6 +1.5 +4.7 +16.5 +41.5 +52.4 +55.8 +0.5 +6.2 1.3 The deviation concerns the difference between the values of the enthalpy of vaporization calculated in this work (Eq. (17)) and the tabulated calorimetric values from the literature. a Calorimetric values from Majer and Svoboda (1985). two cases (n-octane and bromooctane) the agreement is excellent and only in the case of chlorooctane the difference is more important but still considered acceptable. The aqueous solubility of C8H17Cl is circa four times higher than that of n-octane and exhibits a monotonous behavior with temperature in the range studied. This results seems to confirm the tendency observed for the aqueous solubility of chloroalkanes. From the data reported by Horvath and Getzen (1999), the aqueous solubility of 1-chloropentane still exhibits a minimum around 15 C that disappears for temperatures ranging from 3 to 25 C for 1-chlorohexane. No value for the aqueous solubility of chlorooctane was found in the literature for direct comparison. The solubility of C8H17Br in water is higher than that of n-octane (aqueous mole fraction solubilities half of those for C8H17Cl) and exhibits a minimum at lower temperature compared to n-octane. The only value found in the literature for comparison, at 25 C (Horvath and Getzen, 1999), is in reasonable agreement with the result obtained from Eq. (13) with parameters from Table 2 (our value is 9% higher). It is also possible to compare the values of KAW calculated from Eq. (15) at 25 C with the air/water partition coefficients reported for the three solutes in the data base of Yaffe et al. (2003). The literature values of KAW are for n-octane 120, for chlorooctane 1.55 and for bromooctane 2.40. Our values (see below) are about 12% lower for n-octane and for bromooctane and 10% higher for chlorooctane. These differences do not express, however, the quality of our values due to uncertain origin of the literature data. values show large scatter, specially around the minimum of solubility, the overall agreement with our results is satisfactory. We conclude that the experimental values for the aqueous solubility and related coefficients (KH and KAW) exhibit an overall uncertainty below 10%. The equations used for data fitting represent xs, KH and KAW within this error range. According to Eq. (13) with the parameters in Table 2, the aqueous solubility exhibits minimum values for n-octane at 23 C (xs = 1.27 · 107) and for 1-bromooctane at 18 C (xs = 1.57 · 107). It is estimated that the uncertainty associated with the position of the minimum in solubility is of ±1 C, it can be somewhat larger in the case of n-octane for which only scanty data are available at temperatures below 25 C. In the case of chlorooctane, the solubility is four times higher than n-octane and varies monotonously with temperature in the range covered. The values at 25 C are as follows: for n-octane, xs = 1.27 · 107, KH = 147 · 108 Pa and KAW: 107; for bromooctane, xs = 1.70 · 107, KH = 2.97 · 108 Pa, KAW = 2.15; and for chlorooctane xs = 4.67 · 107, KH = 2.35 · 108 Pa and KAW = 1.72. It is believed that the original experimental data reported in this work improve knowledge about the air– water partitioning of halocarbons, substances considered as environmental chemicals. Furthermore a rigorous and compact thermodynamic analysis of the solubility data was performed. It was also shown how these data can be coupled with the vapor pressures of the pure solutes in order to calculate partition coefficients of environmental relevance. 4. Conclusions References Original data for the aqueous solubility are reported for 1-chlorooctane and 1-bromooctane as a function of temperature. These new solubility values were combined with literature data on vapor pressure to determine the HenryÕs law constant and the air/water partition coefficients. The validity of our experimental results and the performance of the experimental technique used were demonstrated by the comparison of the data obtained for n-octane as a function of temperature with the solubility values from the literature. Although the published Abraham, M.H., 1984. Thermodynamics of solution of homologous series of solutes in water. J. Chem. Soc. Faraday Trans. I 80, 153–181. Bergin, G., 2002. Pre´vision de la solubilite´ des hydrocarbures dans lÕeau en fonction de la tempe´rature et de la pression. Universite´ Blaise Pascal, Clermont-Ferrand, France. Boethling, R.S., Mackay, D., 2000. Handbook of Property Estimation Methods for Chemicals—Environmental and Health Sciences. Lewis Publishers, Chelsea, MI, USA. Degranges, S., 1998. Nouvelle procedure de de´termination simultane´e des proprie´te´s enthalpiques et volumiques des S. Sarraute et al. / Chemosphere 57 (2004) 1543–1551 syste`mes fluides: application a` lÕe´tude des solutions aqueuses dÕhydrocarbures jusquÕau domaine critique de lÕeau. Universite´ Blaise Pascal, Clermont-Ferrand, France. DeVoe, H., Miller, M.M., Wasik, S.P., 1981. Generator columns and high pressure liquid chromatography for determining aqueous solubilities and octanol–water partition coefficients of hydrophobic substances. J. Res. Nat. Bur. Stand. 86, 361–366. Dohnal, V., Hovorka, S., 1999. Exponential saturator: a novel gas–liquid partitioning technique for measurement of large limiting activity coefficients. Ind. Eng. Chem. Res. 38, 2036– 2043. Drouillard, K.G., Hiebert, T., Tran, P., Tomy, G.T., Muir, D.C.G., Friesen, K.J., 1998a. Estimating the aqueous solubilities of individual chlorinated n-alkanes (C10–C12) from measurements of chlorinated alkane mixtures. Environ. Toxicol. Chem. 17, 1261–1267. Drouillard, K.G., Tomy, G.T., Muir, D.C.G., Friesen, K.J., 1998b. Volatility of chlorinated n-alkanes (C10–C12): vapor pressures and HenryÕs law constants. Environ. Toxicol. Chem. 17, 1252–1260. Gosset, J.M., 1987. Measurement of HenryÕs law constants for C1 and C2 chlorinated hydrocarbons. Environ. Sci. Technol. 21, 202–208. Heidman, J.L., Tsonopoulos, C., Brady, C.J., Wilson, G.M., 1985. High-temperature mutual solubilities of hydrocarbons and water. Part II: ethylbenzene, ethylcyclohexane, and noctane. AIChE J. 31, 376–383. Horvath, A.L., Getzen, F.W., 1999. IUPAC-NIST Solubility data series 68. Halogenated aliphatic hydrocarbon compounds C3–C14 with water. J. Phys. Chem. Ref. Data 28, 649–777. IUPAC, Commision on Atomic Weights and Isotopic Abundances, 1995. Atomic weights of the elements 1993. J. Phys. Chem. Ref. Data 24, 1561–1576. Jo¨nsson, J.A., Vejrosta, J., Novak, J., 1982. Air/water partition coefficients for normal alkanes (n-pentane to n-nonane). Fluid Phase Equilibria 9, 279–286. Krause, D., Benson, B.B., 1989. The solubility and isotopic fractionation of gases in dilute aqueous solution. IIa. Solubilities of the noble gases. J. Solution Chem. 18, 823–873. Li, J.C., Rossini, F.D., 1961. Vapor pressures and boiling points of the 1-fluoroalkanes, 1-chloroalkanes, 1-bromoalkanes and 1-iodoalkanes, C1 to C20. J. Chem. Eng. Data 6, 268–270. Mackay, D., 1991. Multimedia Environmental Models—The Fugacity Approach. Lewis Publishers, Chelsea, MI, USA. Mackay, D., Shiu, W.Y., 1981. A critical review of HenryÕs law constants for chemicals of environmental concern. J. Phys. Chem. Ref. Data 10, 1175–1199. Majer, V., Svoboda, V., 1985. Enthalpies of Vaporization of Organic Compounds—A Critical Review and Data Compilation. Blackwell, Oxford. May, W.E, Wasik, S.P., Freeman, D.H, 1978. Determination of the Aqueous Solubility of Polynuclear Aromatic Hydrocarbons by a Coupled Column Liquid Chromatographic Technique. Anal. Chem. 50, 175–179. McAuliffe, C., 1963. Solubility in water of C1–C9 hydrocarbons. Nature 200, 1092–1093. 1551 McAuliffe, C., 1966. Solubility in water of paraffin, cycloparaffin, olefin, acetylene, cycloolefine, and aromatic hydrocarbons. J. Phys. Chem. 70, 1267–1275. Miller, D.J., Hawthorne, S.B., 1998. Method for determining the solubilities of hydrophobic organics in subcritical water. Anal. Chem. 70, 1618–1621. Miller, D.J., Hawthorne, S.B., 2000. Solubility of liquid organics of environmental interest in subcritical (hot/liquid) water from 298 K to 473 K. J. Chem. Eng. Data 45, 78–81. Miller, M.M., Wasik, S.P., 1985. Relationships between octanol–water partition coefficient and aqueous solubility. Environ. Sci. Technol. 19, 522–529. Owens, J.W., Wasik, S.P., DeVoe, H., 1986. Aqueous solubilities and enthalpies of solution of n-alkylbenzenes.. J. Chem. Eng. Data 31, 47–51. Polak, J., Lu, C.-Y., 1953. Mutual solubilities of hydrocarbons and water at 0 and 25 C. Can. J. Chem. 51, 4018– 4023. Prausnitz, J.M., Lichetenthaler, R.N., de Azevedo, E.G., 1999. Molecular Thermodynamics of Fluid Phase Equilibria, third ed. Prentice Hall, New Jersey. Price, L.C., 1976. Aqueous solubility of petroleum as applied to its origin and primary migration. Am. Assoc. Petrol. Geol. Bull. 60, 213–244. Reza, J., Trejo, A., Vera-Avila, L.E., 1999. Generator column determination of water solubilities for saturated C6 to C8 hydrocarbons. Int. J. Anal. Chem. 73, 281–295. Ruzicka, K., Majer, V., 1994. Simultaneous treatment of vaporpressures and related thermal data between the triple and normal boiling temperatures for n-alkanes C5–C20. J. Phys. Chem. Ref. Data 23, 1–39. Sandler, S.I., 1999. Chemical Engineering Thermodynamics, third ed. Wiley, New York. Sanemasa, I., Araki, M., Deguchi, T., Nagai, H., 1982. Solubility measurements of benzene and the alkylbenzenes in water by making use of solute vapor. Bull. Chem. Soc. Jpn. 55, 1054–1062. Sedlbauer, J., Bergin, G., Majer, V., 2002. Group contribution method for HenryÕs law constant of aqueous hydrocarbons. AIChE J. 48, 2936–2959. Sijm, D.T.H.M., Sinnige, T.L., 1995. Experimental octanol/ water partition coefficients of chlorinated paraffins. Chemosphere 31, 4427–4435. Smith, J.M., Van Ness, H.C., Abbott, M.M., 1996. Introduction to Chemical Engineering Thermodynamics, fifth ed. McGraw-Hill, New York. Tewari, Y.B., Miller, M.M., Wasik, S.P., Martire, D.E., 1982. Aqueous solubility and octanol/water partition coefficient of organic compounds at 25.0 C. J. Chem. Eng. Data 27, 451– 454. Wagner, W., Pruss, A., 1993. International equations for saturation properties of ordinary water substance. Revised according to the international temperature scale of 1990. J. Phys. Chem. Ref. Data 22, 783–787. Yaffe, D., Cohen, Y., Espinosa, G., Arenas, A., Giralt, F., 2003. A fuzzy ARTMAP-based quantitative structureproperty relationship (QSPR) for the HenryÕs law constant of organic compounds. J. Chem. Inf. Comput. Sci. 43, 85– 112.