Standard partial molal properties of aqueous alkylphenols and alkylanilines over a wide range of temperatures and pressures

Transcription

1 Geochimica et Cosmochimica Acta 71 (2007) Standard artial molal roerties of aqueous alkylhenols and alkylanilines over a wide range of temeratures and ressures Miroslav Čenský a, Josef Šedlbauer b, Vladimir Majer c, *, Vlastimil Růžička a a Deartment of Physical Chemistry, Institute of Chemical Technology, Prague, Czech Reublic b Deartment of Chemistry, Technical University of Liberec, Liberec, Czech Reublic c Laboratoire de Thermodynamique des Solutions et des Polymères, Université Blaise Pascal Clermont-Ferrand/CNRS, F Aubière, France Received 12 May 2006; acceted in revised form 13 October 2006 Abstract This article resents methods for redicting the standard artial molar Gibbs energy (standard chemical otential) and related derivative roerties of aqueous hydroxy and aminoderivatives of (alkyl)benzenes over a wide range of temeratures and ressures. A thorough literature overview was conducted for collecting all available exerimental data resulting from hase equilibrium, calorimetric and volumetric measurements that allow calculation of the thermodynamic roerties of hydration. New exerimental values are resented for solubility in water of isomeric toluidines and for the artial molal volume of henol and cresols at high temeratures. Building uon the acquired database several rediction schemes were develoed and tested for calculating the standard thermodynamic roerties (and namely the Gibbs energy of hydration) of aqueous alkylhenols and alkylanilines as a function of temerature and ressure. First, a simle grou contribution method was roosed for estimations at K and 0.1 MPa using the simultaneous treatment of all available data on hydration roerties at near ambient conditions. Second, this grou contribution method allowed re-adjustment of the arameters of the Helgeson Kirkham Flowers model (HKF) using a new rocedure roosed recently by Plyasunov and Shock [Plyasunov, A.V., Shock, E.L., 2001b. Correlation strategy for determining the arameters of the revised Helgeson Kirkham Flowers model for aqueous nonelectrolytes. Geochim. Cosmochim. Acta 65, ]. Third, using the Sedlbauer O Connell Wood equation of state for aqueous secies (SOCW), grou contributions were determined for redictions at high temeratures and ressures by simultaneous correlation of all available thermodynamic data on hydration roerties. The latter method was constrained by the grou contributions at K and 0.1 MPa making both grou contribution schemes consistent at near ambient conditions. The calculations from the HKF and SOCW equations of state and those from the simle thermodynamic integration of the data at K and 0.1 MPa were comared for several alkylhenols and alkylanilines. Equilibrium constants for hydration reactions obtained from the three aroaches are in very good agreement at temeratures to at least 400 K. At higher temeratures we assess the accuracy of different redictive schemes and their associated uncertainties. The reliable redictions of the standard chemical otentials to at least 573 K and 100 MPa are ossible by the grou contribution method using the SOCW equation of state. Ó 2007 Published by Elsevier Inc. 1. Introduction Hydroxy and aminoderivatives of (alkyl)benzenes (abbreviated here as alkylhenols and alkylanilines) are commonly a art of oilfield waters or waste effluents from various industrial sources, getting in contact with soil and underground waters. Thermodynamic data of these * Corresonding author. Fax: address: vladimir.mayer@univ-bclermont.fr (V. Majer). aqueous solutes are needed at ambient conditions as well as at elevated temeratures and ressures in order to understand or to design the rocesses where the hase and chemical equilibria lay a significant role (Moore et al., 1995; Dale et al., 1997; Taylor et al., 1997; Sheikheldin et al., 2001; Harrison et al., 2002; Feigenbrugel et al., 2004). These are for examle the artitioning of the olar aromatic comounds between the coexisting hydrocarbon-rich hases and dee saline aquifers, the effect of these secies to chemical reactions in the brines coming in /$ - see front matter Ó 2007 Published by Elsevier Inc. doi: /j.gca

2 Proerties of aqueous alkylhenols and alkylanilines 581 contact with rocks, their transfer between the aquatic and air comartments as well as remediation rocesses for removal of the hazardous organic ollutants in soils and underground waters. In this context, the thermodynamic roerties of rime ractical interest for characterizing aqueous solutes are the standard artial molal Gibbs energy of formation, needed in the calculation of chemical equilibria for hydrothermal reactions, and the Gibbs energy of hydration that is closely related to the Henry s law constant imortant in the hase equilibria calculations. There are many tyes of exerimental data that can be used for their evaluation: the equilibrium constants of chemical reactions involving aqueous secies, vaor liquid distribution constants, air water artition coefficients, limiting activity coefficients and solubilities in combination with vaor ressure data. In addition, the data on derivative roerties (enthalies of solution, artial molal heat caacities, and artial molal volumes) resulting from calorimetric and densimetric exeriments are useful in determining the functional deendence of the Gibbs energy on temerature and ressure. Data for henolic comounds were summarized and treated using the HKF thermodynamic model by Dale et al. (1997), allowing redictions of the standard artial molal Gibbs energy of formation as well as the Gibbs henol CH 3 CH 3 CH 3 o-cresol m-cresol -cresol o-dihydroxybenzene m-dihydroxybenzene -dihydroxybenzene NH 2 aniline CH 3 NH 2 CH 3 CH 3 o-toluidine m-toluidine -toluidine NH 2 NH 2 NH 2 NH 2 o-diaminobenzene NH 2 NH 2 NH 2 o-aminofenol m-aminofenol -aminofenol Fig. 1. Molecular structures and the trivial names of rincial solutes treated in this study.

3 582 M. Čenský et al. 71 (2007) energy of hydration in a wide range of conditions. These estimations were based urely on the exerimental data at temeratures below 373 K and at ambient ressure since no other information was available at that time. Large amount of new data became, however, available in the last few years, namely for the derivative roerties (artial molal volumes and heat caacities) of henol, cresols, and dihydroxybenzenes at temeratures to 623 K, allowing a test and refinement of Dale et al. (1997) redictions for henolic solutes. At the same time, analogous data were also obtained for aminoderivatives of aromatic hydrocarbons (aniline, toluidines, diaminobenzenes, and aminohenols), suggesting a ossibility of arallel treatment for this class of solutes. Beside the HKF equation of state (Tanger and Helgeson, 1988), used redominantly by geochemists for the thermodynamic descrition of hydrothermal systems, new models have been ublished recently allowing calculation of the standard thermodynamic roerties of aqueous solutes (e.g., O Connell et al., 1996; Plyasunov et al., 2000a,b; Sedlbauer et al., 2000). A grou contribution concet is often alied for estimating roerties of organic aqueous solutes at the reference conditions of T r = K and r = 0.1 MPa (e.g., Hine and Mookerjee, 1975; Cabani et al., 1981; Meylan and Howard, 1991; Plyasunov and Shock, 2000a, 2001a; Plyasunov et al., 2004, 2005, 2006) or over a wide range of conditions (Amend and Helgeson, 1997; Yezdimer et al., 2000; Plyasunov and Shock, 2000b; Sedlbauer et al., 2002) making these new schemes fully redictive. It is therefore desirable to determine in such new models contributions for hydroxy and amino grous on aromatic rings using the new exerimental data obtained in the last 10 years or so. The urose of this study was: (i) to extend the database for aqueous alkylhenols and alkylanilines by new exerimental measurements, (ii) to collect all thermodynamic data for the given classes of comounds and use them for the develoment of new grou contribution schemes for aqueous alkylhenols and alkylanilines at ambient and at elevated conditions, and (iii) to refine the HKF redictions for aqueous alkylhenols and alkylanilines and comare them with the results of the grou contribution high-temerature model. This article is structured as follows. After a brief introduction to thermodynamic modeling in aqueous solutions of organic comounds we resent in Section 3 new solubility data leading to the Henry s law constant of aqueous toluidines at near ambient conditions and the standard artial molal volumes of henol and cresols at high temeratures. Follows a summary and comments on rimary literature sources that allow the calculation of the thermodynamic roerties of hydration (Gibbs energy, enthaly, heat caacity, and artial molal volume). The grou contributions at the reference conditions of T r = K and r = 0.1 MPa are then determined from the exerimental results at near ambient conditions (T < 373 K). New arameters are listed for the HKF equation of state based on the new data. Parameters of the grou contribution SOCW model are obtained by regression of thermodynamic data at T < 623 K. Finally, the differences between the Gibbs energy results calculated from the different aroaches are quantified and the effect of the derivative roerties of investigated comounds on their high-temerature behavior is discussed. For the sake of clarity the molecular structures of most comounds treated in this study are deicted in Fig Theory The standard 1 artial molal Gibbs energy of formation (or the standard chemical otential for short) of a solute DG f at a temerature T and ressure can be exressed by integration using its known reference value DG f ½T r; r Š and the standard entroy, S [T r, r ] DG f ½T ; PŠ ¼DG f ½T r; r ŠþðT r T ÞS ½T r ; r Š þ Z T T r Z C dt T Z T T r C dlnt þ V d: ð1þ r To evaluate the integrals, exressions for the standard molal volume V = V (T,) and for the standard molal heat caacity C ¼ C ðt ; rþ are needed. Exerimental data and models for these roerties can be thus used for an easy and accurate extraolation of the standard chemical otential at the reference conditions T r and r to high temeratures and ressures. Prime osition among methods of this tye belongs to the revised Helgeson Kirkham Flowers equations (Tanger and Helgeson, 1988) adated to organic aqueous comounds by Shock and Helgeson (1990). New generation of models aeared in the last decade that was insired by the Fluctuation Solution Theory and based on the work of O Connell and collaborators (O Connell et al., 1996; Plyasunov et al., 2000a,b; Sedlbauer et al., 2000). The aroach resented by the last authors (Sedlbauer O Connell Wood) has been alied for data reresentation and for redictions in a variety of aqueous systems, nonelectrolyte and ionic (Majer et al., 2000; Sedlbauer and Majer, 2000, 2004; Yezdimer et al., 2000; Sedlbauer et al., 2002; Sedlbauer and Wood, 2004). Standard chemical otential is the crucial roerty for redicting thermodynamic constants characterizing chemical equilibria in systems involving aqueous solutions. However, for data reresentation and for modeling hase equilibria it is convenient to work in terms of the Gibbs energy of hydration, defined as D h G ¼ DG f ðt ; Þ DGig f ðt ; rþ; ð2þ where DG ig f is the Gibbs energy of formation of ure solute in an ideal gas hase at reference ressure r = 0.1 MPa. 1 The standard state adoted here for aqueous secies is unit activity in a hyothetical one molal solution referenced to infinite dilution.

4 Proerties of aqueous alkylhenols and alkylanilines 583 D h G thus corresonds to the change in free energy associated with the transfer of a solute molecule from an ideal gas hase to an aqueous solution at the standard state of infinite dilution solute ðig;t ; r Þ! soluteðaq;t ; Þ: ð3þ The aroriate equilibrium constant K hyd for this hydration reaction is given by D h G ¼ RT ln K hyd ¼ 2:303RT log K hyd ð4þ The temerature and ressure derivatives of D h G rovide a link with the corresonding standard derivative roerties: enthaly of hydration D h H, heat caacity of hydration D h C, and standard artial molal volume V T 2 oðd h G =T Þ ¼ D h H ; ð5þ ot o ot T 2 oðd hg =T Þ ¼ D h C ot ; ð6þ od h G ¼ V : ð7þ o T Note that in the last relationshi V relaces D h V since the Gibbs energy of an ideal gas in Eqs. (2) and (3) is at constant ressure r. The reason for using a hydration-based reresentation is that most data such as Henry s law constant, limiting activity coefficient, solubility, enthaly of solution, etc. are actually obtained in exeriments close to a hydration rocess; these data are interconnected with the above roerties of hydration by common thermodynamic relations and can be easily converted. The rearrangement to the artial molal roerties of formation (necessary for the G and H functions) requires only ideal gas hase roerties that are well known. Treating just hydration-secific effects also allows develoment of more sensitive and accurate models, as shown, e.g., by Plyasunov et al. (2000a,b) or Sedlbauer et al. (2000). In the case of organic solutes it is often useful to adot a functional grou additivity scheme. The reason is obvious: while the number of organic structures is large, they consist of just a few functional grous. In the thermodynamics of dilute aqueous solutions this rincile has been adoted with success at temerature of 298 K and several methods are available varying in the set of selected functional grous and in the tye of evaluated thermodynamic roerties (e.g., Hine and Mookerjee, 1975; Cabani et al., 1981; Meylan and Howard, 1991; Plyasunov and Shock, 2000a). This aroach can be alied also at elevated temeratures and ressures, where the functional grou contributions need to be exressed in terms of an equation of state such as the HKF model (Amend and Helgeson, 1997) or the SOCW model (Yezdimer et al., 2000; Sedlbauer et al., 2002), or using another emirical reresentation (Plyasunov and Shock, 2000b). Again, the hydration roerties are referred for this treatment because the roerties of ideal gas do not need to be considered in the resulting functional grou contributions. It should also be noted that fundamentally correct grou additivity scheme alies a general equation D h X ¼ X SS þ XN i¼1 n i X i ; where N is the total number of functional grous resent in a given solute, n i is the number of occurrences of each secific functional grou, and X i stands for the X roerty of the ith grou. The term X SS accounts for the intrinsic contribution to the X roerty that is equal to the contribution of hydration of a oint mass. This term is derived from theory (Ben-Naim, 1987) and can be evaluated using only thermodynamic functions of ure solvent (Majer et al., 2004). Functional grous can be reresented by a set of first-order or higher-order contributions (such as steric corrections for roximity effects), deending on the amount of data available for arameterization and the required redictive accuracy of the method. When using a grou contribution concet it is convenient to determine at first the grou contributions at the reference T r, r from the data at near ambient conditions that are relatively abundant and of reasonable accuracy. The arameters of the model reresenting the standard thermodynamic roerties over a wide range of temeratures and ressures can be obtained with the hel of these reference data, as demonstrated for the HKF model by Plyasunov and Shock (2001b). It is of course referable to formulate the model for suerambient conditions again in terms of grou contributions, temerature and ressure deendent this time, which are exressed by a sound, theoretically founded equation. The arameters of these functional deendencies should be determined using simultaneously all the thermodynamic data available in a wide range of conditions. In addition, a rovision should be made to rovide a consistency between the grou contribution schemes for the reference state of T r, r and for suerambient conditions, ideally the two methods should yield identical results at 298 K and 0.1 MPa. This is the aroach we have adoted in this study using the SOCW model (Sedlbauer et al., 2000), resented in Aendix A, in its grou contribution form described by Sedlbauer et al. (2002). 3. Exerimental In this section are resented: (i) new exerimental data for solubilities and the calculated Henry s law constants of o-, m-, and -toluidines determined between 293 and 323 K at the Institute of Chemical Technology Prague, (ii) the standard artial molal volumes of aqueous henol and of o-, m-, and -cresols obtained at 573 and 623 K and at ressures between 10 and 31 MPa from density measurements erformed at the Blaise Pascal University in Clermont-Ferrand. All studied solutes were urchased ð8þ

5 584 M. Čenský et al. 71 (2007) from Fluka in urity of at least 99% in mass and used without further urification Solubility measurements Aqueous solubilities were measured by a conventional batch contacting technique as described by Benes and Dohnal (1999). Aaratus consisted of several jacketed 50 cm 3 glass equilibrium cells connected in series whose temerature was maintained constant to 0.01 K with hel of a Lauda C6 CP (Germany) thermostat. A cooling unit ANKL Kryo (Czech Reublic) was used during measurements at K. Serial connection allowed obtaining simultaneously multile data oints at one temerature; six cells were connected in series for measurement at K while only three cells were used at temeratures of and K. By reducing the number of cells during measurements at higher temeratures we wanted to revent the ossible temerature gradient due to uneven distance of the cells from the thermostat. Outer walls of the cells were covered by cellulose wool for isolation. Studied comounds were added in excess to the cells containing demineralized water and magnetically stirred for aroximately two days. Then the heterogeneous system was allowed to settle under a controlled temerature for aroximately one day. The organic hase was always liquid excet for -toluidine, solid at the lowest exerimental temerature. Samles of the saturated solutions (about 20 cm 3 ) were withdrawn using a syringe through a glass wool filter for analysis. The first ortion of the withdrawn aqueous hase was always discarded in order to avoid ossible adsortion effects on glass wool. Samles were diluted to reach concentration range where the Lambert Beer law is obeyed (absorbance less than 0.8) and analyzed by sectrohotometry using a comuter-interfaced UV absorbance detector LCD 2084 ECOM (Czech Reublic). The concentrations were established on the basis of the revious measurements with three calibration solutions. Two to three solubility determinations were erformed at each temerature with the reroducibility of results being tyically about 1% and lower at the highest temerature. Tables 1a 1c list the results of new measurements in terms of molar fraction x s with their average absolute deviations s x, the literature data are also listed for comarison. In addition, the temerature deendence of solubility data is deicted in Fig. 2. The solubility of o-toluidine ublished by Huyskens et al. (1975) at K was by 4% lower than the value obtained in this work. Chiou et al. (1982) reorted data for o-toluidine and m-toluidine at K; in the case of m-toluidine their value does not seem to be quite consistent with our result at K. The solubility of -toluidine ublished by Hashimoto et al. (1984) at 298 K is by 24% lower in comarison with the value from this study. Our results in combination with the literature values indicate that the solubilities increase with increasing temerature. This is a tye of behavior tyical for hydrohilic solutes unlike Table 1a Solubility of o-toluidine T (K) x sol Æ 10 2 s x Æ 10 2 d sat (MPa) k e H (kpa) a a a b c a This study. b Chiou et al. (1982). c Huyskens et al. (1975). d sat calculated from the correlation of selected exerimental data described below. e Henry s law constant determined as described in the last aragrah of Section Table 1b Solubility of m-toluidine T (K) x sol Æ 10 2 s x Æ 10 2 c sat (MPa) k d H (kpa) a a a b a This study. b Chiou et al. (1982). c sat calculated from the correlation of selected exerimental data described below. d Henry s law constant determined as described in the last aragrah of Section Table 1c Solubility of -toluidine T (K) x sol Æ 10 2 s x Æ 10 2 c sat (MPa) k d H (kpa) a a b a This study. b Hashimoto et al. (1984). c sat calculated from the correlation of selected exerimental data described below. d Henry s law constant determined as described in the last aragrah of Section that for the hydrohobic organic comounds which exhibit a minimum in solubility at a near ambient temerature (tyically between 293 and 323 K). The values for o-toluidine are higher comared to the two other isomers suggesting an increase of solubility due to CH 3 and roximity effect. A different behavior was observed in the case of -toluidine where the solubilities were measured at two temeratures only. The difference between the values at and K is much higher comared to two other isomers. Yet our value at the lower temerature is considerably higher than the literature value (Hashimoto et al., 1984) and the one at the uer temerature is reasonably consistent with the value for m-toluidine. So there is no indication of a ossible systematic error in our measurements and the difference in solubility is aarently

6 Proerties of aqueous alkylhenols and alkylanilines 585 x sol o-toluidine this work (a) 0.20 (b) m-toluidine this work (a) 0.15 (b) -toluidine this work (c) due to the fact that -toluidine is solid at the lower temerature and subcooled liquid at the uer one. It is aarent from the text below (see Section 4.1.1) that while the Henry s law constant and the activity coefficient are a continuous function of temerature indeendent of the hysical state of ure solute, the solubility is strongly affected by the solute s state. Tables 1a 1c list also for the three toluidines the values of vaor ressures and of the resulting Henry s law constants k H obtained by the rocedure described below (see Section 4.1.1) Volumetric measurements T (K) Fig. 2. Exerimental data on solubility of the three toluidines: ref a, Chiou et al. (1982); ref b, Huyskens et al. (1975); ref c, Hashimoto et al. (1984). The standard artial molal volumes were evaluated from the density differences between aqueous solutions of henol or cresols and water measured on a vibrating tube flow densimeter (Hynek et al., 1997). The densities are obtained from the oscillation eriods of a U tube, vibrating in a field of a ermanent magnet. Since this is a comarative method a calibration exeriment with two fluids of wellknown density is necessary. The temerature within one exeriment was stable to 0.01 K and was measured with an accuracy of 0.03 K using a secondary standard latinum thermometer (Burns, 500 X). Pressure was maintained stable to ±0.02 MPa by means of a back-ressure regulator (Circle Seal) and read by means of an electronic manometer (DPI 260 Druck) with an exected accuracy of 0.05 MPa. The difference between the densities of a solution q and deionized water q w is calculated as Dq ¼ q q w ¼ Kðs 2 s 2 w Þ; ð9þ where s and s w are the eriods of vibration of the tube filled successively with solution and water. Calibration constant K was obtained by measurements with water and nitrogen. Densities were obtained at target temeratures of 573, 598, and 623 K and at a ressure close to the saturation ressure of water. Determinations at the two higher temeratures were also erformed at a ressure near 30 MPa. Measurements were carried out at u to five concentrations between 0.15 and 0.75 m for henol and at two target concentrations (0.1 and 0.2 m) for cresols, the uer concentration limit being a function of the solubility. Two to three exeriments were erformed with each solution at a given temerature and ressure. The measured density differences were converted to the aarent molar volumes (Majer et al., 2004) V / ¼ M s =ðq w þ DqÞ Dq=ðmq w ðq w þ DqÞÞ; ð10þ where M s and m are molar mass of water and molality of the solute, resectively. The extraolation to zero concentration allows to obtain the artial molar volume of solute at infinite dilution that has the meaning of the standard artial molal volume V ¼ lim m!0 V / ¼ M s =q w lim m!0 ðdq=mþ=q 2 w : ð11þ In the case of henol the limiting value of the ratio Dq/m was obtained by fitting the exerimental data to a linear relationshi Dq=m ¼ a þ bm ð12þ using the weighted least-squares regression. The reroducibility of measurements and the exected error in the calibration constants served for determining the weighting factors. In the case of cresols, for which the concentration range is limited, it was not ossible to determine the intercet a with a good accuracy. It was referred to take as the V value the arithmetic average of the aarent molar volumes that were obtained at sufficiently low concentrations to allow an aroximation of infinite dilution. The exected errors in V were 2 3 cm 3 mol 1 for henol and somewhat higher for cresols (3 4 cm 3 mol 1 ). The differences between volumes for the three isomers were lower than the exected uncertainty of the standard artial molal volumes. For that reason, we resent only the averaged values considered as reresentative for all three cresols. These values are listed in Table 2 together with three literature values available close to our exerimental conditions and resulting from highly reliable sources (listed below Table 2). The volumetric roerties of henol and the three cresols are available from Hynek et al. (1997) and from Hnedkovsky et al. (1998) at the lower temerature limit of our measurements; the differences with our values are below 1 cm 3 mol 1, which confirms our error estimation. Criss and Wood (1996) reorted the standard artial molal volume of henol at 598 K and 28 MPa that is 2.4 cm 3 mol 1 higher than our value measured at 31.4 MPa. Since the standard artial molal volume of aqueous nonelectrolytes generally decreases with increasing ressure the two values can be considered as consistent. New results and the values determined earlier at lower temeratures at the Institute of Chemical Technology in Prague are lotted in Fig. 3 that illustrates the reasonable agreement of results from

7 586 M. Čenský et al. 71 (2007) Table 2 Standard artial molal volumes of aqueous henol and cresols at high temeratures and ressures, new measurements and literature data T (K) (MPa) V (henol) (cm 3 mol 1 ) V (cresols) a (cm 3 mol 1 ) b c d a Averaged value for o-, m-, and -cresols. b Hynek et al. (1997). c Degrange (1998) reorted 172.4, 172.7, and cm 3 mol 1 for o-, m-, and -cresols, resectively. d Criss and Wood (1996). V o (cm 3.mol -1 ) henol (a) this work (a) (31.0 MPa) this work (31.4 MPa) cresols (b) this work this work (31.0 MPa) T (K) Fig. 3. Standard artial molal volumes as a function of temerature (T > 400 K) and ressure. The values were obtained close above the saturation ressure of water when not otherwise indicated: ref a, Hynek et al. (1997); ref b, Hnedkovsky et al. (1998). different data sources as a function of temerature and ressure. 4. Thermodynamic roerties of aqueous alkylhenols and alkylanilines 4.1. Review of exerimental data The sources of exerimental data directly related to the Gibbs energy of hydration and its temerature and ressure derivatives for the studied classes of solutes are listed in Tables 3 6. These sources were used to constitute a database of 160, 17, 156, and 264 data oints leading, resectively, to D h G, D h H, D h C, and V values used in the simultaneous correlation. Beside the solution roerties, the data on ure solutes are also needed for conversion between the solution and hydration roerties. This is namely the case of vaor ressures, enthalies of vaorization and sublimation and ideal gas heat caacities whose sources are also secified below. It is aarent that a considerable amount of data is available on the Gibbs energy level at temeratures below 373 K resulting from a variety of literature sources. On the other hand, only few enthalic values were reorted for a limited number of solutes at or near 298 K. The volumetric and heat caacity data are available for both hydroxy and aminoderivatives over a wide range of temeratures and ressures thanks to recent camaigns of measurements at the Prague Institute of Chemical Technology (Czech reublic) and at the Blaise Pascal University Clermont-Ferrand (France). Phenol, cresols, and aniline are the solutes for which the highest number of data is available in the literature. It can be noted that for solutes containing two olar grous (dihydroxybenzenes, diaminobenzenes, and aminohenols) no data are available for the Gibbs energy and enthaly while sufficient information is available for the volumes and heat caacities Gibbs energy of hydration The Gibbs energy of hydration is directly related to the Henry s law constant k H D h G ¼ RT lnðk H = r ÞþRT lnðm w m Þ ð13þ in which r = 0.1 MPa, m o = 1 mol kg 1, and k H is defined as the limiting ratio of the fugacity to the mole fraction of a solute in an aqueous hase: k H ¼ limðf =xþ: ð14þ x!0 The term containing the molar mass of water M w is the conversion term between the molar fraction and molality standard state conventions. The Henry s law constant can be obtained for the studied class of solutes from different tyes of exerimental data characterizing dilute aqueous solutions at near ambient conditions. It is simly related to the vaor liquid distribution constant (equal to the limiting value of the relative volatility) K d ¼ lim x!0 ðy=xþ where y and x are the molar fractions of a solute in coexisting vaor and liquid hases, resectively. At near ambient conditions k sat,w K d where sat,w is the vaor ressure of water. Another source is the air water artition coefficient K aw ¼ lim x!0 ðc a =C w Þ where C stands for molarity. This coefficient is used in environmental chemistry for exressing the artitioning of a ollutant between atmosheric and aquatic hase; it is called sometimes the dimensionless Henry s

8 Proerties of aqueous alkylhenols and alkylanilines 587 Table 3 Data sources leading to the Gibbs energy of hydration, D h G c Solute Tem. range a Press. range b No. of data Reference Phenol Schreinemakers (1900) Weller et al. (1963) Hakuta (1975) Leunberger et al. (1985) Abd-El-Bary et al. (1986) Trem et al. (1993) Dohnal and Fenclova (1995) Moore et al. (1995) Tabai et al. (1997) Sheikheldin et al. (2001) Harrison et al. (2002) Feigenbrugel et al. (2004) o-cresol Hakuta (1975) Leunberger et al. (1985) Trem et al. (1993) Dohnal and Fenclova (1995) Sheikheldin et al. (2001) Harrison et al. (2002) Feigenbrugel et al. (2004) m-cresol Leunberger et al. (1985) Dohnal and Fenclova (1995) Sheikheldin et al. (2001) Feigenbrugel et al. (2004) -Cresol Leunberger et al. (1985) Trem et al. (1993) Dohnal and Fenclova (1995) Feigenbrugel et al. (2004) 2,3-Dimethylhenol Leunberger et al. (1985) Dohnal and Fenclova (1995) Sheikheldin et al. (2001) 2,4-Dimethylhenol Leunberger et al. (1985) Dohnal and Fenclova (1995) Sheikheldin et al. (2001) 2,5-Dimethylhenol Leunberger et al. (1985) Dohnal and Fenclova (1995) 2,6-Dimethylhenol Leunberger et al. (1985) Dohnal and Fenclova (1995) 3,4-Dimethylhenol Leunberger et al. (1985) Dohnal and Fenclova (1995) 3,5-Dimethylhenol Leunberger et al. (1985) Dohnal and Fenclova (1995) Aniline Dallos et al. (1983) Moore et al. (1995) Bernauer et al. (2006) o-toluidine Chiou et al. (1982) This work m-toluidine Chiou et al. (1982) This work -Toluidine Moore et al. (1995) This work a K. MPa. c Additional sources with only one result at 298 K and 0.1 MPa: Parsons et al. (1971) for henol; Parsons et al. (1972) for o-, -cresol; Huyskens et al. (1975) for o-toluidine; Hashimoto et al. (1984) for -toluidine; Jayasinghe et al. (1992) for aniline, -toluidine, 3,4-dimethylaniline, and 2,4, 5-trimethylaniline; Shiu et al. (1994) for 2,4,6-trimethylhenol, 2,4-dimethylhenol, 2,6-dimethylhenol, 4-nonylhenol, 4-octylhenol, henol and o-, m-, and -cresol; Mackay et al. (1995) for 2- and 4-ethylhenol; Altschuh et al. (1999) for o-, m-cresol, aniline and o-, m-, and -toluidine. law constant. It simly stands k H = K aw RT/V w (V w is the molar volume of water). Values for the limiting activity coefficients c Rµ comlying with Raoult s law and derived from vaor liquid, chromatograhic or solubility exeriments have been ublished in chemical engineering and environmental literature at temeratures u to 373 K or so. They can be converted to the Henry s law constant by multilication with vaor ressure of ure liquid solute (real or hyothetical) according to k H ¼ l sat cr1. For solids or liquids saringly soluble in water, the Henry s law

9 588 M. Čenský et al. 71 (2007) Table 4 Data sources leading to the enthaly of hydration, D h H c Solute Tem. range a Press. range b No. of data Reference Phenol Fernandez and Heler (1959) Parsons et al. (1971) Nichols and Wadsö (1975) Gillet (1990) o-cresol Parsons et al. (1972) -Cresol Parsons et al. (1972) Nichols and Wadsö (1975) Aniline Nichols and Wadsö (1975) Gillet (1990) a K. b MPa. c The enthalies of sublimation of henol, o-cresol, and -cresol were obtained from Andon et al. (1960), the enthaly of vaorization of subcooled -cresol was obtained from Andon et al. (1960), the enthaly of vaorization of aniline was from the review of Majer and Svoboda (1985), resectively. Table 5 Data sources for the standard artial molal volume, V c Solute Tem. range a Press. range b No. of data Reference Phenol Criss and Wood (1996) Hynek et al. (1997) Hnedkovsky et al. (1998) Origlia-Luster et al. (2003) o-cresol Hnedkovsky et al. (1998) Censky et al. (2005a) m-cresol Hnedkovsky et al. (1998) Censky et al. (2005a) -Cresol Makhatadze et al. (1990) Hnedkovsky et al. (1998) Censky et al. (2005a) o-dihydroxybenzene Jedelsky et al. (1999) m-dihydroxybenzene Jedelsky et al. (1999) -Dihydroxybenzene Jedelsky et al. (1999) Aniline Ruzicka et al. (2000a) o-toluidine Ruzicka et al. (2000b) m-toluidine Ruzicka et al. (2000b) -Toluidine Ruzicka et al. (2000b) o-diaminobenzene Hyncica et al. (2002) m-aminohenol Striteska et al. (2003) a K. b MPa. c Additional sources with only one result at 298 K and 0.1 MPa: Desnoyers et al. (1973), Hamann and Linton (1974), Hamann and Lim (1954), and Hokins et al. (1976) for henol; Indelli (1963) for o-, m-, and -dihydroxybenzene; Shahidi et al. (1977) for aniline. constant can be also aroximated by the ratio of the vaor ressure of ure solute and its solubility in water k H = sat /x sol where x sol is the mole fraction solubility. This aroach is not, however, easily alicable here since most solutes are fairly soluble and the vaor ressures at near ambient conditions are often missing for these comounds of low volatility. It should be noted that beside the rigorous thermodynamic definition of k H by Eq. (14) that we strictly adot here, other definitions are used in the literature. The Henry s law constant is considered in geochemistry and atmosheric chemistry rather as a concentration to ressure ratio and it is usually assimilated to K aw by environmental chemists. All these different data allowing the calculation of D h G via k H (Eq. (13)) can be found in references listed in Table 3. It is aarent that more information is available for the hydroxyderivatives, where most of data were reorted for henol and cresols, than for solutes containing the amino grou. It is ractically imossible to derive k H values for solutes with two olar grous on the aromatic ring that are highly soluble in water and nonvolatile (solid) at ambient conditions. For the solutes containing one olar grou, Henry s law constants are ublished or can be derived from the reorted data at 298 K or over a limited temerature interval close to and above ambient conditions. The data have been ublished in sources focusing on thermodynamic data for chemical engineering and in journals of environmental and atmosheric chemistry. On the thermodynamic side, the most imortant contribution is that of Dohnal and Fenclova (1995) who ublished highly reliable Henry s law constants and limiting

10 Proerties of aqueous alkylhenols and alkylanilines 589 Table 6 Data sources leading to heat caacity of hydration D h C c Solute Tem. range a Press. range b No. of data Reference Phenol Origlia-Luster et al. (2003) Censky et al. (2005a) o-cresol Censky et al. (2005a) m-cresol Censky et al. (2005a) -Cresol Makhatadze et al. (1990) Censky et al. (2005a) o-dihydroxybenzene Censky et al. (2005b) m-dihydroxybenzene Censky et al. (2005b) -Dihydroxybenzene Censky et al. (2005b) Aniline Censky et al. (2005a) o-toluidine Censky et al. (2005a) m-toluidine Censky et al. (2005a) -Toluidine Censky et al. (2005a) o-diaminobenzene Censky et al. (2005b) o-aminohenol Censky et al. (2005b) m-aminohenol Censky et al. (2005b) -Aminohenol Censky et al. (2005b) a K. b MPa. c Additional sources with only one result at 298 K and 0.1 MPa: Perron and Desnoyers (1979) and Hokins et al. (1976) for henol; Bernauer et al. (2006) for aniline. Table 7 Values of grou contributions for calculating thermodynamic roerties of hydration of aqueous alkylhenols and alkylanilines at 298 K and 0.1 MPa Grou D h G a D h H a D h C b V c C d CH d d CH d CH d,e C ar d,e CH ar I(C C) f n.e. n.e. g,h hi 19.11(0.13) j 27.51(0.45) j 30(6) j 12.88(0.36) j g,h NH 2,hi 15.96(0.20) j 26.43(0.75) j 49(8) j 16.71(0.72) j I(C ) f,h 1.83(0.18) j n.e. k 24(24) j 0.8(0.8) j I( ) f,h n.e. k n.e. k 51(30) j 2.2(1.4) j I(NH 2 NH 2 ) f,h n.e. k n.e. k 69(32) j 3.0(2.2) j i X SS a kj mol 1. b JK 1 mol 1. c cm 3 mol 1. d Plyasunov and Shock (2000a). e A grou with subscrit ar is a art of an aromatic ring. f Correction for ortho-osition on the aromatic ring: C stands for alkyl-, for hydroxy-, and NH 2 for amino grous. g A grou with subscrit hi is directly bound to aromatic ring. h This work. i Standard state term (see Eq. (8)). j 95% confidence limits of the arameter. k Not evaluated (no data available or arameters are not significant at the 95% confidence level). activity coefficients determined in a vaor liquid equilibrium circulation still for 10 henolic solutes. Bernauer et al. (2006) recently reorted highly reliable k H values for aniline that are the result of a simultaneous treatment of new vaor liquid equilibrium and calorimetric data. Other values for henol were measured by Abd-El-Bary et al. (1986) using the inert gas striing method and by Tabai et al. (1997) using ebulliometry. In the latter reference are reorted both the Henry s law constants and limiting activity coefficients. These data, however, are not consistent due to a calculation error in the original aer when determining k H that we have corrected. A similar aroach was used by Moore et al. (1995) who reorted limiting activity coefficients and the Henry s law constants for henol, aniline, o-toluidine, and -toluidine determined by differential ebulliometry. However, the data for o-toluidine which seem to be in error were not considered. Dallos et al. (1983) ublished one limiting activity coefficient for aniline. All conversions between c Rµ and k H were erformed by us using the vaor ressures reorted by Dohnal and Fenclova (1995) for henolic comounds and by Censky (2001) for solutes with the amino grou. The vaor ressure data for this latter grou of comounds were obtained by combining the new static measurements (Censky, 2001) with the literature data (Dreisbach and Schrader, 1949; McDonald et al., 1959; Krevor et al., 1985; Steele et al., 1994). They were correlated by the Cox equation as a function of temerature simultaneously with the ideal gas and liquid heat caacities (Frenkel et al., 1994; Zabransky et al., 1996) in order to rovide a realistic extraolation of vaor ressures towards ambient temeratures. This tye of data treatment was described earlier by Ruzicka and Majer (1996). A myriad of aroaches was used by environmental and atmosheric chemists in determinations of data leading to k H at ambient conditions. Some of these data are, however, of low reliability and striking differences are observed between the values originating from different sources. This has been recently documented in an article by Harrison et al. (2002), in which an overview table lists the Henry s

11 590 M. Čenský et al. 71 (2007) Table 8 Gibbs energy of hydration, D h G, and Gibbs energy of formation in the ideal gas, DG ig f, of aqueous alkylhenols and alkylanilines at K and 0.1 MPa Solute DG ig a f D h G a GC value a,b Phenol 32.5 c 19.78, d 19.98, e 15.97, f 18.72, g h o-cresol 34.3 c 15.97, g 16.08, h 16.68, i 17.24, e 15.07, f j m-cresol 40.1 c 16.59, f 17.44, g 17.00, h j Cresol 31.5 c 17.44, f 17.25, g 18.22, h i Ethylhenol 25.7 r k Ethylhenol 25.7 r k Octylhenol 24.8 r h Nonylhenol 33.2 r h ,3-Dimethylhenol 33.2 c g ,4-Dimethylhenol 41.1 c 14.83, g h ,6-Dimethylhenol 38.9 c 12.77, g h ,4-Dimethylhenol 34.1 c g ,5-Dimethylhenol 39.3 c g ,4,6-Trimethylhenol r h Aniline 7.0 c 15.52, j 15.42, l m o-toluidine c 15.43, j 14.81, n o m-toluidine c 15.86, j 14.9 n Toluidine c 17.81, j,q 15.14, l 14.39, n ,4-Dimethylaniline r l ,4,5-Trimethylaniline r l a kj mol 1. b Grou contribution value. c Frenkel et al. (1994). d Parsons et al. (1971). e Harrison et al. (2002). f Feigenbrugel et al. (2004). g Leunberger et al. (1985). h Shiu et al. (1994). i Parsons et al. (1972). j Altschuh et al. (1999). k Mackay et al. (1995). l Jayasinghe et al. (1992). m Bernauer et al. (2006). n This work. o Huyskens et al. (1975). Hashimoto et al. (1984). q Excluded from evaluation. r Estimated from Joback s grou contribution scheme (Poling et al., 2001). Table 9 Enthaly of hydration, D h H, and enthaly of formation in the ideal gas, DH ig f, of aqueous alkylhenols and alkylanilines at K and 0.1 MPa Solute DH ig a f D h H a GC value a,b Phenol 96.4 c 55.94, d 55.96, e 55.69, f h o-cresol c g Cresol c 59.79, g h Aniline 87.5 c 53.93, e h a kj mol 1. b Grou contribution value. c Frenkel et al. (1994). d Parsons et al. (1971). e Gillet (1990). f Fernandez and Heler (1959). g Parsons et al. (1972). h Nichols and Wadsö (1975). law constants of henol at 298 K ublished in literature. The highest value differs from the lowest by more than one order of magnitude. In addition, some more reliable data from thermodynamic sources were ignored. We have extracted the Henry s law constants from the data obtained recently using different vaor liquid equilibrium techniques for several henolic comounds (Sheikheldin et al., 2001; Harrison et al., 2002; Feigenbrugel et al., 2004). A considerable amount of k H data was obtained at 298 K only (the references are listed at the bottom of Table 3), articularly for solutes containing NH 2 grou (Jayasinghe et al., 1992; Altschuh et al., 1999). One should also mention solubility measurements by Chiou et al. (1982) and Hashimoto et al. (1984) for aniline and toluidines (see also Tables 1a 1c) that we have converted to k H using vaor ressures of Censky (2001) as indicated above. In Table 3 are also listed, beside rimary data sources, three review articles reorting k H values whose origin is sometimes unclear. This is the case of Henry s law constants of henolic solutes obtained at 281 and 298 K as a ratio of different vaor ressure and solubility data by

12 Proerties of aqueous alkylhenols and alkylanilines 591 Table 10 Standard artial molal volume, V, of aqueous alkylhenols and alkylanilines at K and 0.1 MPa Solute V a GC value a,b Phenol 86.2, c 86.1, d 86.17, e 86.06, f 86.0 g o-cresol e m-cresol e Cresol 100.3, h e o-dihydroxybenzene 87.07, i j m-dihydroxybenzene 88.92, i 88.8 j Dihydroxybenzene 88.7, i 88.7 j Aniline 89.3, k l o-toluidine m m-toluidine m Toluidine m o-diaminobenzene n m-aminohenol o a cm 3 mol 1. b Grou contribution value. c Hokins et al. (1976). d Hamann and Linton (1974). e Hnedkovsky et al. (1998). f Desnoyers et al. (1973). g Hamann and Lim (1954). h Makhatadze et al. (1990). i Indelli (1963). j Jedelsky et al. (1999). k Shahidi et al. (1977). l Ruzicka et al. (2000a). m Ruzicka et al. (2000b). n Hyncica et al. (2002). o Striteska et al. (2003). Leunberger et al. (1985). Several values at 298 K were taken from the reviews of Shiu et al. (1994) and Mackay et al. (1995). The only source reorting directly D h G (298 K) determined from solubilities and vaor ressures for henol, o-cresol and -cresol is Parsons et al. (1971, 1972). In order to increase the number of D h G data for toluidines, we have determined the Henry s law constants at three temeratures by combining the results of solubility measurements resented above with vaor ressures of ure solutes. Since toluidines are only moderately hydrohobic and hence relatively soluble, we did not aroximate k H as the vaor ressure to solubility ratio but referred to convert first the solubility to limiting activity coefficient c Rµ comlying with Raoult s law. While the relationshi is trivial between solubility and activity coefficient in the case of liquid solutes (c R =1/x sol ), conversion is more comlex in the case of -toluidine that is solid at the lowest temerature: c R ¼ F =x sol F ¼ s sat =l sat ¼ exðd msð1 T m =T ÞÞ; ð15þ where the entroy of melting D m S (54.22 J K 1 mol 1 )at the melting oint temerature T m = 317 K was determined by Censky et al. (2001). Once the activity coefficient is determined, its limiting value is obtained by aroximating its concentration deendence using the Margules equation as roosed by Wright et al. (1992) Table 11 Heat caacity of hydration, D h C, and heat caacity of the ideal gas, Cig, of aqueous alkylhenols and alkylanilines at K and 0.1 MPa Solute C ig a D h C a 220 Phenol c 212.9, d 210.5, e 222.1, f 217.9, g o-cresol c f 278 m-cresol c f 254 -Cresol c 275.8, f h 254 o-dihydroxybenzene c i 203 m-dihydroxybenzene c i 152 -Dihydroxybenzene c i 152 Aniline c 229.1, f j 239 o-toluidine c f 273 m-toluidine c f 273 -Toluidine c f 273 o-diaminobenzene k i 258 -Diaminobenzene k i 189 o-aminohenol k i 171 m-aminohenol k i 171 -Aminohenol k i 171 GC value a,b a JK 1 mol 1. b Grou contribution value. c Frenkel et al. (1994). d Perron and Desnoyers (1979). e Hokins et al. (1976). f Censky et al. (2005a), at K. g Origlia-Luster et al. (2003) at 0.35 MPa. h Makhatadze and Privalov (1990). i Censky et al. (2005b), at K. j Bernauer et al. (2006). k Estimated from Joback s grou contribution scheme (Poling et al., 2001). ln c R1 ¼ ln c R ð1 x sol Þ 2 : ð16þ Then the Henry s law constant was calculated by multilying the c Rµ value by the vaor ressure of liquid toluidines determined by Censky (2001). In all calculations of the limiting activity coefficients for toluidines, it was assumed that their activity in the organic hase is unity. Desite the nonnegligible solubility of water in the liquid organic hase, generally higher than that of organic solute in water, this aroximation is generally considered as reasonable at near ambient conditions. This is due to the fact that the activity coefficient (comlying with the Raoult s law) of a hydrohobic organic comound saturated by water is always slightly higher than unity comensating thus aroximately solubility of water in the organic hase Enthaly of hydration The enthaly of hydration is obtained by combining the enthalies of solution D s H obtained from calorimetric dissolution exeriments with the enthalies of vaorization or sublimation D v H of ure liquid or solid solutes: D h H ¼ D s H D v H : ð17þ Both D s H and D h H are ublished in literature and the inconsistency between different sources can be due to the different D v H values. For that reason we have used systematically as inut exclusively the calorimetric enthalies