Prediction and experimental verification of the effect of salt on the vapour–liquid equilibrium of ethanol/1-propanol/water mixture

https://doi.org/10.1016/j.fluid.2005.05.019Get rights and content

Abstract

The vapour–liquid equilibria of ethanol/1-propanol/water mixture saturated with sodium chloride, potassium chloride and copper sulphate predicted by Tan–Wilson and Tan–NRTL phase models compared well with the experimental data. The prediction was based on the solvent–solvent interaction parameters obtained from the regression of the experimental vapour–liquid equilibrium of the salt-free mixture and the solute–solvent interaction parameters calculated from the bubble points of the individual solvent components saturated with the respective salt. The extent of salting-out of the solvent components from the liquid phase by the three salts generally decreased in the order of 1-propanol, ethanol and water giving a vapour phase rich in 1-propanol and ethanol and a liquid phase rich in water at equilibrium. These effects are slightly different between sodium and potassium chloride and are significantly greater than those of copper sulphate. These observed salting effects are consistent with the criteria that solvent component, i, would be preferentially salted-in or out of the liquid phase than solvent component, j, if the ratio of their solute–solvent interaction parameters Asj/Asi (Tan–Wilson) or exp(τis  τjs) (Tan–NRTL) is less or greater than one. These results confirm the good accuracy of the two phase models in describing the effect of a non-volatile solute on the liquid phase activity of a solvent component in a ternary solvent mixture. These results suggest that the rapid screening method and criteria for selecting a suitable non-volatile solute to separate close-boiling and azeotropic binary solvent mixture developed by Tan based on these phase models can be extended to include the ternary solvent mixtures.

Introduction

A soluble non-volatile solute (electrolyte or non-electrolyte) in a solvent mixture would generally affect the distribution of the solvent components between the equilibrium phases. This results from the difference in the interaction forces between the dissolved solute and the various solvent components, thereby modifying their liquid phase activity differently. This phenomenon has been exploited for the distillation of close-boiling and azeotropic aqueous–solvent mixtures with the deliberate addition of a suitable dissolved inorganic salt. However, it is not as commonly used as the extractive and azeotropic distillation processes which use an extraneous solvent to accentuate the deviation of the relative volatility of the solvent components in the mixture away from unity. Although the salting phenomena would similarly affect the liquid–liquid equilibrium of a solvent mixture, its application to solvent–solvent extraction of solutropic or near solutropic mixtures has not been reported.

Research has been focused mainly on the experimental determination and correlation of the vapour–liquid equilibria (VLE) of binary aqueous–solvent mixtures containing an inorganic salt and considerably less or none relating to non-volatile non-electrolytes, non-aqueous solvent mixtures, ternary and other multicomponent solvent mixtures and liquid–liquid equilibrium (LLE) of solvent–solute mixtures. Furter and co-workers [1], [2], [3], [4], [5], [6] correlated the VLE of binary aqueous–solvent mixtures containing a dissolved salt with the Wilson model [7] using salt-free liquid composition and activity coefficients of the solvent components based on their saturation vapour pressure modified for the presence of the salt. Ohe and co-workers [8], [9], [10], [11] advocated modifying the liquid composition with the solvation parameters of the solvent components while using the unmodified saturation vapour pressure of the solvent components. Other modifications of the liquid composition have also been reported [12], [13]. Chen and co-workers [14], [15], [16] described the salt effects in terms of the interaction between each salt ion and each of the solvent components. Similar approach based on group contribution was also adopted by other workers [17], [18], [19], [20], [21], [22] using phase models such as UNIF AC [23] and UNIQUAC [24]. The pertinent salt effect parameters in all these methods, including the modified UNIF AC and UNIQUAC models, were obtained from the regression of the experimental VLE data of the mixture according to the respective model. Tan [25] described the liquid phase activity of the solvent component in a solvent–solute (electrolyte and non-electrolyte) mixture on the local volume composition basis (similar to the Wilson model for salt-free solvent mixture) and derived a phase model described by two sets of interaction parameters, namely the solvent–solvent and the solute–solvent interaction parameters. The set of solvent–solvent interaction parameters is identical with the set of Wilson's parameters for solute-free solvent mixture and is obtained by the regression of the experimental VLE data of solute-free mixture. The set of the solute–solvent interaction parameters is calculated from the bubble points of the individual solvent components containing the same solute concentration as in the mixture. With these two sets of interaction parameters, the VLE of the solvent–solute mixture can be predicted without relying on its experimental VLE data. Similar to Wilson model, this local volume composition model is not able to describe liquid phase separation. Adopting the same conceptual approach, Tan [26] derived a local molar composition model to include liquid phase separation similar to the NRTL model [27]. Unlike Chen and co-workers [14], [15], [16], Tan in his models [25], [26] considered the total net interaction effect between each solvent component and the solutes without detailing out the individual interaction between each solvent components and each of the solutes and their dissociated, ionic and undissociated products. This concept was experimentally verified by Tan and Ng [28] with the good prediction of the VLE of a binary aqueous–solvent mixture containing two dissolved salts. This conceptual approach significantly reduces the number of solute–solvent interaction parameters compared with the approach which detailed out the individual interaction between each solvent component and each of the ions of the dissolved salt [14], [15], [16], [17], [18], [19], [20], [21], [22]. Tan et al. [29] found that non-volatile non-electrolytes, such as glucose showed similar’ salting phenomena’ and the VLE was similarly well predicted by both the models. Hence, it was found appropriate to replace “salt” with “non-volatile solute” to include the non-electrolytes. Both the models developed by Tan [25], [26] predicted reasonably accurately the VLE of binary solvent mixtures containing single and two solutes (electrolytes and non-electrolytes) [25], [26], [30], [31], [32] and the VLE of ternary solvent mixture of water/ethanol/2-propanol containing sodium chloride, potassium chloride, sodium nitrate or potassium acetate [33]. Tan found in all these cases [25], [26], [30], [31], [32], [33], that the solvent component i is preferentially salting-in or out of the liquid phase relative to solvent component j in a solvent–solute mixture would depend on whether the ratio of the solute–solvent interaction parameters, Asj/Asi (Tan–Wilson model) or exp(τis  τjs) (Tan–NRTL model) is less or greater than 1. Tan [34], [35] developed a rapid screening technique for a suitable solute for the salt (solute) distillation of close-boiling and azeotropic mixtures and the criteria for the elimination of azeotrope by the dissolved solute(s). This paper reports the comparison of the experimental VLE data of ethanol/1-propanol/water saturated with sodium chloride, potassium chloride and copper sulphate with those predicted by both the models proposed by Tan.

Section snippets

Materials and analytical method

Analytical grade chemical reagents from Baker or Merck and filtered deionised water from a Millipore water purification system were used to determine the VLE of ethanol (1)/1-propanol (2)/water (3) mixture with and without sodium chloride, potassium chloride or copper sulphate at 760 mmHg. The salts were pre-dried for at least 6 h at 200 °C in an oven and cooled in a dessicator before they were used. Solvent mixtures were analyzed with a Shimadzu GC-17A gas chromatograph using a Chromosorb 102

Results and discussion

Table 1 gives the experimental VLE of the ethanol (1)/1-propanol (2)/water (3) mixture at 760 mmHg and the corresponding Wilson and NRTL solvent–solvent interaction parameters and NRTL non-randomness factors obtained from the correlation of the experimental activity coefficient of the solvent components according to the respective models. The saturation vapour pressures of the solvent components for the calculation of their activity coefficients were calculated according to the Antoine equation

Conclusion

The vapour–liquid equilibrium of ethanol/1-propanol/water was modified by dissolved sodium chloride, potassium chloride and copper sulphate to give a vapour phase enriched in 1-propanol and ethanol and lean in water. The extent of salting-out of the solvent components decreased in the order of 1-propanol, ethanol and water. The salting effects of sodium chloride and potassium chloride differ slightly but are significantly greater than those of copper sulphate. This may be attributed to the

References (37)

  • S. Ohe

    Fluid Phase Equilib.

    (1998)
  • Y.L. Gao et al.

    Fluid Phase Equilib.

    (1991)
  • Y.L. Gao et al.

    Fluid Phase Equilib.

    (1991)
  • B. Sander et al.

    Chem. Eng. Sci.

    (1986)
  • E.A. Macedo et al.

    Chem. Eng. Sci.

    (1990)
  • I. Kikic et al.

    Chem. Eng. Sci.

    (1991)
  • T.C. Tan et al.

    Fluid Phase Equilib.

    (1988)
  • T.C. Tan et al.

    Fluid Phase Equilib.

    (2004)
  • T.C. Tan

    Fluid Phase Equilib.

    (1990)
  • W.F. Furter, Salt Effect in Vapor–liquid Equilibria, Ph.D. Thesis, University of Toronto,...
  • A.I. Johnson et al.

    Can. J. Chem. Eng.

    (1960)
  • D. Meranda et al.

    AIChE J.

    (1971)
  • D. Jaques et al.

    AIChE J.

    (1972)
  • D. Meranda et al.

    AIChE J.

    (1972)
  • D. Meranda et al.

    AIChE J.

    (1974)
  • G.M. Wilson

    J. Am. Chem. Soc.

    (1964)
  • S. Ohe et al.

    IHI Eng. Rev.

    (1969)
  • S. Ohe

    Adv. Chem. Ser.

    (1976)
  • Cited by (17)

    • Accelerating flash calculation through deep learning methods

      2019, Journal of Computational Physics
      Citation Excerpt :

      Knowledge of the equilibrium conditions in mixtures can be obtained from data collected in direct experiment or using thermodynamic models, including activity coefficients at low system pressure or fugacity coefficients at high system pressure [22]. In the last two decades, the application of equilibrium calculation using equations of state (EOS) that describes mixing rules with experience coefficients has been proposed and widely discussed [9,29,30]. A realistic EOS, e.g., Peng-Robinson (PR EOS), is generally considered as an appropriate thermodynamic model to correlate and predict vapor–liquid equilibrium (VLE) conditions, due to the long-time improvement developed with applications in different aspects.

    • New computational tool to evaluate experimental VLE and VLLE data of multicomponent systems

      2017, Computers and Chemical Engineering
      Citation Excerpt :

      This is an example of a quaternary system formed by a mixture of three alcohols in water. The thermodynamic model required to pass the consistency test is constructed using binary (Kurihara et al., 1993; Orchillés et al., 2008; Lladosa et al., 2008a,b) and ternary VLE data (Lladosa et al., 2006; Tan et al., 2005), in addition to hE (Dohnal and Rehak, 2007; Hovorka et al., 2002; Nagamachi and Francesconi, 2006; Pope et al., 1967; Ramalho and Ruel, 1968). The integral-form of the test verifies all the experimental data, Fig. 13a, reflecting that the relationship between the compositions and temperature has an acceptable consistency.

    • Isobaric vapor-liquid-liquid-solid equilibrium of the water + NaCl + 1-butanol system at 101.3 kPa

      2016, Journal of Chemical Thermodynamics
      Citation Excerpt :

      However, experimental data are needed to check the predictions and correlations of the models. Regarding the water + alcohol + salt systems, the experimental data available in literature does not cover all of the regions [6–12]. In fact, it is difficult to find data of solid–liquid–liquid–vapor because the different authors tend to avoid the equilibrium determination of the liquid–liquid region.

    • Salting-out effect of ionic liquids on isobaric vapor-liquid equilibrium of acetonitrile-water system

      2015, Chinese Journal of Chemical Engineering
      Citation Excerpt :

      When ILs dissociate into anions and cations, the hydration between these ions and water molecules takes place due to electrostatic interactions, the result of which is that the number of free water molecules as solvent is reduced, the relative volatility of acetonitrile to water is improved, and the separation of materials is promoted. Tthe salt effect theory is often applied to predict vapor-liquid equilibrium with inorganic salt solution, and the prediction is acceptable [14–17]. Literally ionic liquids are a kind of molten salt with similar nature as inorganic salt solution.

    • Double-effect integration of multicomponent alcoholic distillation columns

      2012, Energy
      Citation Excerpt :

      The same criterion of the binary parameters evaluation (ΔyA < 0.03 and ΔyR < 10%) was used in this case. VLE (Vapor-liquid equilibrium) experimental data were found [32,37–51] for 38 ternary systems and for 4 quaternary systems involving some components (ethanol, water and 9 congeners) present in the wine. The total number of experimental data was defined as (C – 1) ∙ S, where C is the number of components in each system (ternary or quaternary) and S, the total number of systems.

    View all citing articles on Scopus
    View full text