Determination of infinite dilution activity coefficients using HS-SPME/GC/FID for hydrocarbons in furfural at temperatures of (298.15, 308.15, and 318.15) K

https://doi.org/10.1016/j.jct.2012.01.016Get rights and content

Abstract

A new methodology using the headspace solid phase microextraction (HS-SPME) technique has been used to evaluate the infinite dilution activity coefficient (γ12) of nine hydrocarbons (alkanes, cycloalkanes, and aromatics) in furfural solvent. The main objective of this study was to validate a faster and lower cost methodology expanding the use of HS-SPME to determine infinite dilution activity of solutes in organic solvents. Two approaches were proposed for the determination of γ12 in order to use this technique (HS-SPME). In addition, the fiber–gas partition coefficients (Kfg) for each analyte at each of the studied temperatures were determined. The activity and partition coefficients have been reported at temperatures of (298.15, 308.15, and 318.15) K. The data were compared with the literature infinite dilution data determined by other methods such as liquid–gas chromatography (GLC) and gas stripping. Partial molar excess enthalpies of mixing at infinite dilution for each solute have been determined. The fibers were tested before and after each experiment, using statistical methods to ensure that their properties do not change during the experiments. The fibers were also analyzed by optical microscopy to evaluate possible surface damage by comparing them with new fibers. The activity coefficient values correlated well with the data in the literature and showed average deviations less than 10%.

Highlights

► Two approaches were proposed using SPME on determination of infinite dilution activity coefficients. ► Infinite dilution activity coefficients of nine solutes in solvent furfural at T = (298.15, 308.15, and 318.15) K. ► Fiber–gas partition coefficients of nine solutes on PDMS at T = (298.15, 308.15, and 318.15) K. ► Optical microscopy analysis and statistical tests to measure possible damages on fiber coating. ► Advantages and limitations of methodology proposed were discussed.

Introduction

At infinite dilution, a molecule of a solute is completely surrounded by solvent molecules. In this conformation, the maximum non-ideality is generally indicated and the behavior of the solute is defined by the infinite dilution activity coefficient (γ12) [1], [2]. The knowledge of γ12 allows the determination of various thermodynamic parameters of industrial and theoretical importance [3], [4]. Knowing the γ12 of each component in the other in a binary mixture allows the determination of binary interaction parameters of activity coefficient models and adjustment of the binary interaction parameters in equations of state [3], [4]. These parameters can be used to make predictions of (vapor + liquid) equilibrium over all ranges of compositions in a given temperature. The importance of activity coefficients at infinite dilution in the industry is reflected in projects involving thermal separation processes such as liquid–liquid extractions, azeotropic distillation, stripping process, and other types of processes where high purity products are necessary [5], [6], [7]. An example of this application is in the use of furfural, an extraction solvent and a byproduct from sugarcane bagasse. In the petroleum industry, furfural is used in the extraction of aromatic compounds from lubricating oils. The presence of aromatics can cause changes in viscosity when the final product is exposed to temperature variation. The knowledge of γ12 is necessary for the complete removal of these undesired compounds [8], [9].

Many techniques are described in the literature for the determination of γ12, however, all the methods present limitations that prevent their use in various types of systems. Gas–liquid chromatography (GLC) is the most used of these techniques and is ideal for the determination of γ12 in non-volatile or low volatile solvents [1], [10]. Currently, most of the activity coefficient data in the literature are reported for ionic liquids through the GLC technique because the results are precise and relatively easy to obtain [11], [12], [13], [14]. The disadvantage of this technique lies in the difficulty or impossibility to determine both limiting values of infinite dilution activity coefficients of binary mixtures [1], [10]. Other techniques used are inert gas stripping and differential ebulliometry [15], [16]. The inert gas stripping presents the key advantage of determination of infinite dilution activity coefficients for many compounds on a single experiment; however, this technique is only suitable for systems with high relative volatility. On the other hand, the differential ebulliometry technique is suitable for determination of limiting values of activity coefficients of both components in a binary mixture; nevertheless, the experimental apparatus is difficult to operate for most of the highly non ideal systems [1], [10], [15], [16].

The most common method to measure infinite dilution activity coefficients for polymeric systems is frequently called as the gas chromatography method, inverse gas chromatography, or gas–liquid chromatography [17], [18], [19], [20]. This method uses the partition coefficients or specific retention volumes of solute (VN) to determine the activity coefficient. The material to be analyzed (polymeric stationary phase) is coated inside a gas chromatography column. The retention times of solute and a non-sorbed compound in an isothermal run are measured, allowing the determination of the partition coefficient and specific retention volume between the solute and polymeric phase [17], [20]. There are other methods to determine partition coefficients [21], [22], [23], [24], [25]. One of these methods uses the Linear Temperature Programmed Retention Index (LTPRI) approach. The LTPRI technique allows the determination of many partition coefficients on a single chromatographic run. The logarithms of the partition coefficient of homologous series of hydrocarbons are linearly related to its LTPRI number (by definition, 100 times the number of carbon atoms). For any other compound, only the LTPRI number is necessary to estimate the partition coefficients using the linear regression of the curve that correlates the logarithm of partition coefficients and LTPRI number [22].

Solid phase microextraction (SPME) is a technique of extraction and concentration of analytes for posterior analysis on analytical equipment such as gas chromatography (GC) or high performance liquid chromatography (HPLC). SPME uses a fine silica rod, 110 μm long, coated with a polymeric coating to extract and concentrate target analytes present in a matrix. The critical component of SPME fibers is the material of the coatings. These coatings have specific characteristics, like thickness and polarity, that directly influences the kinetics of extraction [26], [27].

In the past few years, researchers showed that SPME can be used to determine partition and activity coefficients in polymeric systems and activity coefficients at infinite dilution of solutes in non-polymeric and non-organic liquid solvents such as water [28], [29]. Other studies tried to extent the application of this technique in high concentrated and complex matrices by evaluating possible changes in the sorption characteristics of the SPME fibers [30], [31].

In this study, in a continuation of a previous project [2], the goal is to extent the headspace SPME (HS-SPME) for the determination of infinite dilution activity coefficients of solutes in organic solvents with accuracy, easiness, and lower costs than the conventional methods. Two approaches are presented here to determine the activity coefficients in organic solvents using SPME. The values of γ12 and partial molar excess enthalpies at infinite dilution of nine hydrocarbons in furfural were determined and compared with literature data. Statistical tests and optical microscopy analysis were performed to assure reproducibility and measure possible structural damages to fibers caused by high swelling rates. Characteristics of the technique such as advantages, limitations, and result deviations from data reported in the literature were also evaluated.

Section snippets

Theoretical basis

The definition of a closed system containing a liquid phase, a gaseous phase, and the polymeric phase of the SPME fiber is necessary for the determination of the infinite dilution activity coefficient by headspace SPME (HS-SPME) as showed in figure 1.

The definition of a closed system implies that the total amount of solute will be distributed between the gas, liquid, and polymeric phases. It means that the total mass of solute (n1T) is equal to the initial mass of solute (n10)dn10dt=dn1Tdt=0n10

Materials

The chemicals used in this study were purchased from Vetec. The chemicals used as solutes were n-pentane, n-hexane, n-heptane, cyclopentane, cyclohexane, benzene, toluene, ethylbenzene, and o-xylene. The solvent used was furfural; and methanol was used for the construction of calibration curves. All chemicals were analyzed by gas chromatography and refractometry; all purities were better than provided as shown in table 1. The chemicals were used without further purification. SPME fibers with 30 

Results and discussion

The calibration curves were determined for all chemicals, except for methanol. The determination coefficient (R2) for all calibration curves ranged from 0.9976 for furfural to 0.9995 for n-hexane.

The extraction time for the 30 μm PDMS fibers was determined for each solute and were all less than 1 min. To ensure the equilibrium of solute with the fiber coating, the extraction time was assumed equal to 4 min. The desorption time was assumed as 4 min. After each extraction, the fibers were re-exposed

Conclusion

Two approaches were proposed to extent the SPME in the determination of activity coefficients at infinite dilution for solutes in organic solvents. The technique proved useful in most of the studied cases using the two proposed approaches. Average deviations of approximately 10% were observed in all tested systems and dependent on the approach used. Approach 1 produced better results than Approach 2 for compounds with liquid–gas partition coefficients lower than fiber–gas partition

References (42)

  • J. Bao et al.

    Fluid Phase Equilib.

    (1995)
  • S.I. Sandler

    Fluid Phase Equilib.

    (1996)
  • B. Coto et al.

    Chem. Eng. Sci.

    (2006)
  • K. Kojima et al.

    Fluid Phase Equilib.

    (1997)
  • P. Yan et al.

    J. Chem. Thermodyn.

    (2010)
  • P. Reddy et al.

    J. Chem. Thermodyn.

    (2011)
  • A. Marciniak

    J. Chem. Thermodyn.

    (2011)
  • J.D. Raal et al.

    Fluid Phase Equilib.

    (2001)
  • P. Haimi et al.

    Fluid Phase Equilib.

    (2006)
  • A. Voelkel et al.

    J. Chromatogr. A

    (2009)
  • J.M. Santiuste

    Anal. Chim. Acta

    (2001)
  • M. Dudek et al.

    J. Chromatogr., A

    (2002)
  • V. Dohnal et al.

    Fluid Phase Equilib.

    (2010)
  • A. Jahnke et al.

    J. Chromatogr. A

    (2010)
  • A. Kloskowski et al.

    J. Chem. Thermodyn.

    (2005)
  • E. Santacesaria

    Fluid Phase Equilib.

    (1979)
  • T.M. Letcher et al.

    J. Chem. Thermodyn.

    (2004)
  • F.A. Furtado et al.

    Quim. Nova

    (2010)
  • C.S. Schacht et al.

    Ind. Eng. Chem. Res.

    (2010)
  • U. Domańska et al.

    J. Phys. Chem. B

    (2010)
  • J.C. Bastos et al.

    Ind. Eng. Chem. Process. Des. Dev.

    (1985)
  • Cited by (0)

    View full text