Experimental Solid–Liquid Phase Equilibria of a Methyl Ester/Amide/Nitrile Ternary System by DSC A. Mekki-Berrada, S. Bennici, J.L. Dubois & A. Auroux
Journal of the American Oil Chemists' Society ISSN 0003-021X Volume 90 Number 11 J Am Oil Chem Soc (2013) 90:1621-1627 DOI 10.1007/s11746-013-2320-2
1 23
Your article is protected by copyright and all rights are held exclusively by AOCS. This eoffprint is for personal use only and shall not be self-archived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”.
1 23
Author's personal copy J Am Oil Chem Soc (2013) 90:1621–1627 DOI 10.1007/s11746-013-2320-2
ORIGINAL PAPER
Experimental Solid–Liquid Phase Equilibria of a Methyl Ester/ Amide/Nitrile Ternary System by DSC A. Mekki-Berrada • S. Bennici J.-L. Dubois • A. Auroux
•
Received: 29 January 2013 / Revised: 27 March 2013 / Accepted: 22 July 2013 / Published online: 18 August 2013 Ó AOCS 2013
Abstract Binary and ternary mixtures of lauric methyl ester, amide and nitrile were investigated by differential scanning calorimetry, in order to determine the solid–liquid phase equilibria of this model for saturated organic systems. Eutectic mixtures were observed and followed in the ternary system, while solid–solid transitions and immiscibility phenomena were also characterized. This system is also of industrial interest for processes in the frame of biomass valorization. Keywords Fatty nitrile Ternary organic system Differential scanning calorimetry
Introduction Matsuoka and Ozawa [1] pointed out how the number of organic materials being separated and/or purified by crystallization in industrial scale operations is increasing rapidly (low energy consumption and possibility of obtaining products with a higher purity than by other conventional separation operations). However, modeling of organic solid– liquid equilibria is not as well developed as for metal oxides or salts, or as well as liquid–gas equilibria, although some recent literature can be found [2–8]; thus accurate empirical data has to be sought. Whereas several data banks or tools concerning liquid–gas equilibria are available for organic A. Mekki-Berrada S. Bennici A. Auroux (&) IRCELYON, UMR5256 CNRS-Universite´ Lyon1, 2 avenue A. Einstein, 69626 Villeurbanne Cedex, France e-mail:
[email protected] J.-L. Dubois ARKEMA, Direction Recherche and De´veloppement, 420 Rue d’Estienne d’Orves, 92705 Colombes, France
mixtures, it is quite difficult to find data about liquid–solid equilibria. An index of lipid phase diagrams was documented by Koynova and Caffrey [9], however it is mainly focused on lipids of membrane origin where water is the dispersing medium. Fatty esters and their mixtures as organic phase change materials are also recommended as energy storage materials due to their desirable thermal and heat transfer characteristics and the advantage of easy impregnation into conventional building materials [10], thus several mixtures of esters were studied, however not with fatty amides and/or nitriles. This article focuses on an organic ternary mixture of lauric methyl ester, amide and nitrile, for which no similar literature can be found. In the frame of valorization of biomass, the conversion of triglycerides into fatty acids or esters and then into fatty nitriles (Fig. 1) is of industrial interest, either as a platform molecule for polymer chemistry or as a fraction of a biofuel blend [11]. Then the fatty amide intermediary product becomes a key aspect of developed processes, since its melting point is by far higher than both starting material and final product. This issue has to be controlled both during the process and the final state of purification; this is where knowledge on the physical properties of the ester/ amide/nitrile ternary becomes necessary. Techniques were developed to remove impurities from nitriles, and for example in the case of low contents of amide, using mineral acids in order to protonate the diluted amide and bring it to precipitate and be removed mechanically. Frank et al. [12] claimed in this process the removal of the diluted content of amide from industrially produced nitrile, up to the saturation point about 0.9 wt% at room temperature, but this most probably refers to the sharp increase in the melting temperature of the mixture with increasing amide content for low values, which will be discussed and quantified in this article.
123
Author's personal copy 1622
Fig. 1 Simple scheme of the conversion of lauric methyl ester into nitrile
Some attention also has to be given to the stability of the studied mixtures, this is the absence of chemical reaction during the differential scanning calorimetry (DSC) experiment. Without a catalyst and at temperatures lower than 423 K, the dehydration of amide towards nitrile is very slow [13] and the reaction that can proceed between an acid and a nitrile [14] is less probable by far when it concerns the methyl ester and the nitrile.
Experimental Procedure Materials Lauric methyl ester (C98 %, Sigma-Aldrich), lauramide [[96 %(T), TCI] and lauronitrile [C98 %(GC), TCI] were used in this study. GC–MS analyses (Perkin-Elmer Clarus 580-560 S, Elite WAX ETR 30/0.25/0.5 column) did not let us identify any major impurity. Samples of various compositions were prepared and melted into miscible solutions, from which about 5 mg was put into sealed aluminum DSC crucibles. Relative statistic error on weighing of the filled DSC crucibles was about 4 %. Mixtures of above 1 g were prepared using a precision balance (M-220, Denver Instruments, precision 10-4 g), ensuring good precision on the composition in ester/amide/nitrile.
J Am Oil Chem Soc (2013) 90:1621–1627
been used for investigating fatty acid methyl esters’ melting behavior, with a similar apparatus, also for pure esters and binary or ternary mixtures [15]. Measured cooling ramps here provide more separated peaks, however severe undercooling can be observed, and then only heating ramps are considered for matters of phase transition temperature and enthalpy, although the cooling ramps provide interesting qualitative information. Temperature limits of domains can be determined from the heating DSC curves by the following method: a peak corresponding to an invariant reaction can be treated as the melting peak of a pure compound and the onset temperature has to be considered; in the case of diphasic domains, however, the composition changes with temperature resulting in a variation of thermal capacity and thus of the baseline of the DSC curve, then the liquidus limit temperature is determined by considering the extremum of the curve and correcting it with the slope assigned to the solidus peak (see Fig. 2) [16]. The mass of the samples could not be reduced enough to separate peaks, and changing the heating rate was observed to not separate peaks when decreasing from 5 to 1 K/min. Besides, no transition appeared to be missed by heating at 5 K/min instead of 1 K/min. During heating and cooling of these mixtures inside the 1 g mother solutions, no immiscibility could be observed in the liquid phase. Besides, mixtures containing more than 1 mol% amide were observed to solidify into gels, denoting some homogeneous solidification with trapped liquid species, whereas at about 1 mol% of amide, some small solid particles were observed to appear inside the liquid volume.
Results and Discussion Procedure The DSC apparatus was a Q100 model from TA instruments, with precise control over both heating and cooling programs via a cold and hot source system. Enthalpy and heat capacity calibration were performed with indium (melting point at 429.75 K) and sapphire respectively. Deionized water (272.8 ± 0.2 K and 322.8 ± 12.8 J/g measured, instead of 273.16 K and 333 J/g) melting temperature and enthalpy were also measured and compared to the literature, denoting a slight deviation in the reading of measured temperature lower than 0.4 K. Besides, the evaluation of enthalpies is mostly hindered by the precision of the weighing and was used only qualitatively. The temperature program consisted of an equilibration at 423 K, in order to ensure the melting and homogeneity of samples, followed by a cooling ramp at 5 K/min until 183 K, then a heating ramp at 5 K/min back to 413 K. Cooling ramps and cooling/heating ramps have already
123
Solid–Liquid Equilibria for the Binary System of Methyl Laurate and Lauronitrile Methyl laurate and lauronitrile are aprotic molecules and melt at similar temperatures around 276–278 K [17–20].
Fig. 2 a DSC peak for an invariant reaction, with onset temperature T0, and b DSC peak corresponding to diphasic domains and the calculated temperature T0’ of the domains’ limit
Author's personal copy J Am Oil Chem Soc (2013) 90:1621–1627
Fig. 3 Heating DSC curves for the lauric ester/nitrile binary system; ester content in mol% is increasing from top to bottom
Fig. 4 Derived solidus (black squares), liquidus (blue triangles) and solvus (red circles) from the ester/nitrile binary system; domains’ temperature limits are plotted as a function of the ester content (mol%)
The melting point observed in this study for pure methyl laurate was at 276.5 ± 0.5 K, and for pure lauronitrile about 274.6 ± 0.5 K. Deviation from the literature data could be ascribed to purity matters. Measured melting enthalpies for the ester and the nitrile are respectively 36.5 ± 3.7 kJ/mol (lit.: 43.10 kJ/mol [20]) and 30.7 ± 3.0 kJ/mol (lit. estimation: 28.34 kJ/mol [17]). The DSC curves for the methyl laurate/lauronitrile binary system are plotted in Fig. 3, while derived solidus, liquidus and solvus points can be observed on the binary in Fig. 4. Superposition of DSC peaks below 273 K is quasi systematic and hypotheses on the nature (solidus/liquidus) of the peaks and the shape of the solidus can help with deriving accurate domain limits. A small peak about 263 K can be observed on every mixture (not decipherable in Fig. 3 with the present scale for the sake of clarity of the series of peaks) and can be assigned to an immiscibility domain in-between lauric ester and nitrile, since its
1623
temperature tends to decrease in the proximity of both pure compounds, while being rather stable otherwise, perhaps due to the proximity of the solid–liquid equilibrium curve. The presence of a eutectic mixture can be observed at 264 K and 35 ± 1 mol% of methyl laurate. The solidus curve displays a plateau at 263.5 K over a large content from about 15 to 55 mol% of ester. The binary diagram reveals the miscibility of both solids above 263 K and the shape of domains of ester enriched and nitrile enriched solids can be observed on the left and right of the diagram respectively, while in between the solidus and liquidus domains of miscible liquid ? enriched solid can be observed on the left and right of the eutectic point. Both molecules being similar by their saturated fatty chain and the aproticity and dipolar moment of their function, as well as by their melting point, such a diagram seems coherent. Both molecules can weakly interact either by their fatty saturated chains or by the dipolar moments of their functions. Steric effects could however explain their limited miscibility in solid phase, or also the probably different crystallization structure, since the nitrile function is linear, whereas the methyl ester function is planar and asymmetric. The methyl stearate’s crystal structure is reported to be monoclinic [21], and methyl laurate being a similar saturated fatty methyl ester probably crystallizes as a similar structure. Butyronitrile is reported to crystallize in space group P21/a, however its chain is far shorter than the presently studied nitrile [22]. Some cocrystallization however seems to happen in the temperature frame above 263 K, and the observed slightly endothermic process that we assign to mixing corresponds to an enthalpy of about 1 J/g. Other organic binaries were observed to demix below their melting point [23], and the entropy difference ends up overcoming the enthalpic stabilization of some interaction (here eventually dipole1-dipole2). Demixing enthalpy for PEO/PES polymer blends was observed as a slight enthalpic effect inside the measured heat capacity by MTDSC, and was evaluated below 5 J/g [24]. Enthalpic analysis of this binary can be observed in Fig. 5, where the enthalpic contributions of solidus and liquidus points are displayed as a function of the ester content. The very low contribution of the solidus for ester content above 70 mol% confirms the location where the solidus leaves the eutectic temperature. Transition enthalpy at the eutectic is about 34.7 kJ/mol, in-between pure ester and nitrile enthalpies. Solid–Liquid Equilibria for the Binary Systems of Lauramide with Either Methyl Laurate or Lauronitrile Lauramide is a protic molecule and melts at a temperature about 369 to 373 K [25], which we observed at
123
Author's personal copy 1624
Fig. 5 Enthalpic contributions of the liquidus (black squares) and the solidus (red circles) in the ester-nitrile binary
373.8 ± 0.2 K by DSC. It was reported to present a solid– solid transition at about 322 K (and enthalpy 9.4 kJ/mol), which we also observed at 321.8 K ± 0.2 (and enthalpy 8.8 ± 1.0 kJ/mol) for all mixtures containing more than 1 mol% of amide. Lauramide solidifies in the form of white crystals with centrosymmetric hydrogen-bonded pairs (space group P21/a) [25]. Since several hydrogen bonds are available for organizing the solid structure, it can be proposed that the solid–solid transition might involve organization relative to weaker interactions such as the hydrophobic saturated carbon chains. The lauramide/methyl laurate binary system was studied by DSC, and analyses of various compositions are displayed in Fig. 6. A liquidus curve can be observed at the highest temperatures between pure solid amide and the liquid phase, decreasing from 374 K for pure amide to about 333 K for 4 mol% of amide. Then the solid–solid equilibrium of amide can be observed at 322 K. By continuity with the 1 mol% amide isoplete (this is the section of the ternary diagram where amide content is constant, here at the 1 mol% value), the solidus is observed around 310 K, below the solid–solid transition of lauramide, and the mixture appears as a gel. A similar shape can be observed for the solidus of the methyl ester, stable at 276.5 K from pure ester until about 10 mol% of ester, where it starts decreasing. It can be observed that both molecules are immiscible in their solid form, while they are miscible in liquid phase. The influence of the mixture on the melting temperature is only appreciable at high content of ester or nitrile, and the solid–solid equilibrium of amide remains unchanged by the presence of methyl ester. The lauramide/lauronitrile binary system was studied by DSC, and analyses of various compositions are displayed on Fig. 7. A solidus curve can be observed at highest
123
J Am Oil Chem Soc (2013) 90:1621–1627
Fig. 6 Derived domains’ temperature limits for the ester/amide binary system, as a function of the ester content (mol%)
Fig. 7 Derived domains’ temperature limits for the nitrile/amide binary system, as a function of the nitrile content (mol%)
temperatures between pure solid amide and the liquid phase, decreasing from 374 K for pure amide to about 323 K for 3 mol% of amide. Then the solid–solid equilibrium of amide can be observed at 322 K. By continuity with the 1 mol% amide isoplete, the solidus is observed at 308 K, below the solid–solid transition of lauramide, and the mixture appears as a gel. A similar shape can be observed for the solidus of the nitrile, decreasing slowly from 274.5 K for pure nitrile until 272 K for about 5 mol% of nitrile. It can be observed that both molecules are immiscible in their solid form, while they are miscible in liquid phase. The influence of the mixture on the melting temperature is only appreciable at high content of the counter molecule, and the solid–solid equilibrium of amide remains unchanged by the presence of nitrile.
Author's personal copy J Am Oil Chem Soc (2013) 90:1621–1627
1625
(mol%)
(mol%) (mol%) Fig. 8 Ternary coordinates of the isopletes for 1, 15, 30, 50 and 70 mol% amide
Solid–Liquid Equilibria for the Ternary System of Lauramide, Methyl Laurate and Lauronitrile In order to appreciate the shape of the ternary system, several isopletes were performed, and behaviors of the amide with either the nitrile or the ester being similar, it was decided to focus on isopletes of the amide, this is while keeping the content of amide constant while varying the relative amounts of both other molecules. Five series of experiments were performed with amide contents of 0.95 ± 0.1, 14 ± 1.5, 28 ± 1, 50 ± 0.5 and 70 ± 1 mol% of amide, and their ternary coordinates can be seen in Fig. 8. Shapes observed on every isoplete are similar to the binary ester/nitrile, with enriched solids, a eutectic and an immiscibility lacuna, which means that the amide does not interact much with the binary, at least until 70 mol% of amide. It can be figured that amide is either nucleating and crystallizing or reticulating the system into smaller volumes of identical ester/nitrile composition. Besides, the mixture’s melting point is determined by amide and is almost constant on each amide isoplete. Figure 9 displays these melting points as a function of amide content alone. In Fig. 10 are displayed the temperature limits of domains as a function of the relative ester to nitrile content for these five isopletes, focused on the lower temperature part below 322 K, where the amide has already solidified. Then there is no eutectic, since the amide brings the system into a solid or gel state, however, the ester/nitrile mixture appears to keep properties close to its eutectic shape discussed in the previous part, thus from here on we will call the extension of the ester/nitrile eutectic point in the ternary mixture a ‘‘pseudo-eutectic’’. First, it can be observed that the location of this ‘‘pseudo-eutectic’’ is rather stable in terms of ester/nitrile composition, although some deviation could be occurring, as depicted on Fig. 11. First at
Fig. 9 Melting point (black squares) and solid–solid transition (red circles) for the amide isopletes
35 ± 1 mol% of ester relative to nitrile for the mixture without amide, it then tends to 37 ± 1 mol% at 1 mol% of amide, then it appears around 41 ± 2 mol% at 15 mol% of amide, around 39 ± 2 mol% at 30 mol% of amide, around 40 ± 4 mol% at 50 mol% of amide, and finally around 40 ± 4 mol% (of ester relative to nitrile) at 70 mol% of amide. The standard error on the overall composition stays about 1.5 % and the tendency for high amide content is stable around 39 mol% of ester relative to nitrile, higher than for the binary. The temperature of the ‘‘pseudo-eutectic’’ mixture seems to vary slightly from 264 to 266 K between the pure binary and the eutectic at 1 mol% of amide; however it is stable at 265 K for higher amide content. Second, the immiscibility lacuna also remains unchanged with a culminating point at 263 K and a large plateau. It was reported that amide only dissolves until about 1 mol% into nitrile [12]. If extrapolated to the ester/nitrile mixture, this could explain the more similar shape between every isoplete above 1 mol% of amide, than with the pure binary system, the latter having larger domains of solid ? liquid, while the former display a thinner shape, there is a shorter temperature difference between solidus and liquidus (of the ester-nitrile domains). This would imply the interaction of the dissolved 1 mol% of amide with both ester and nitrile, which would shift upwards the ‘‘solidus’’. In conclusion, the investigated methyl laurate/lauronitrile binary system is found to display a eutectic diagram with a eutectic point at 265 K and a composition of 35 ± 1 mol % of ester, which does not seem much affected by the addition of amide, with which it seems immiscible. A shift of the ‘‘pseudo-eutectic’’ mixture however appears until a composition of 39 mol% of ester (relative to nitrile) at high amide content, and the shape of the esternitrile diagram appears thinner when amide is in the mixture. Besides, lauramide displays immiscibility with both
123
Author's personal copy 1626
J Am Oil Chem Soc (2013) 90:1621–1627 b Fig. 10 Solid-liquid equilibria for the isopletes with 1, 15, 30, 50 and
70 mol% amide as a function of the ester content relative to both ester and nitrile, focus on low temperature range
Fig. 11 Ternary coordinates estimation of the eutectic mixtures for the amide isopletes (between red circles and black squares)
other compounds and their mixtures, and the melting point of these mixtures decreases at low amide content until temperatures below the solid–solid transition (322 K) for amide concentrations lower than 1 mol%. This denotes the industrial difficulty of handling aprotic mixtures with a high content of amide, for the melting point will not be much lowered. Amide’s miscibility with the ester/nitrile mixture being lower than 1 mol%, mechanical/thermal processes can be used to remove this fraction, and furthermore the already discussed patent [12] proposes a process to remove the remaining diluted content. Methyl laurate and lauronitrile are miscible in the liquid phase and their boiling points are closer than 10 K at atmospheric pressure, thus separation by distillation appears difficult and chemical treatment would probably be needed for separation. Acknowledgments The authors are thankful to the scientific services of IRCELYON. The research leading to these results has received funding from the European Union Seventh Framework Programme (FP7/2007-2013) under grant agreement no. 241718 EuroBioRef.
References 1. Matsuoka M, Ozawa R (1989) Determination of solid-liquid phase equilibria of binary organic systems by differential scanning calorimetry. J Cryst Growth 96:596–604 2. Domanska U, Marciniak M (2007) Experimental (solid ? liquid) or (liquid ? liquid) phase equilibria of (amine ? nitrile) binary mixtures. J Chem Thermodyn 39:247–253
123
Author's personal copy J Am Oil Chem Soc (2013) 90:1621–1627 3. Witusiewicz VT, Hecht U, Sturz L, Rex S (2006) Phase equilibria and eutectic growth in ternary organic system (D)camphor–neopentylglycol–succinonitrile. J Cryst Growth 286:431–439 4. Shibuya H, Suzuki Y, Yamaguchi K, Arai K, Saito S (1993) Measurement and prediction of solid-liquid phase equilibria of organic compound mixtures. Fluid Phase Equilib 82:397–405 5. Letcher TM, Redhi GG (2002) Phase equilibria for liquid mixtures of (butanenitrile ? a carboxylic acid ? water) at 298.15 K. Fluid Phase Equilib 193:123–133 6. Ozawa R, Matsuoka M (1989) Determination of solid-liquid phase equilibrium of organic ternary eutectic mixtures by differential scanning calorimeter—the o-, m- and p-nitroaniline system. J Cryst Growth 98:411–419 7. Desseyn HO, Clou K, Janssens JF, Carleer R (2003) Vibrational and structural evidence of phase transitions in some amides. Thermochim Acta 402:81–90 8. Gonzalez JA, Garcia de la Fuente I, Cobos JC, Casanova C (1994) Thermodynamics of mixtures containing linear monocarboxylic acids. I. DISQUAC predictions on molar excess Gibbs energies, molar excess enthalpies and solid- liquid equilibria for mixtures of linear monocarboxylic acids with organic solvents. Fluid Phase Equilib 99:19–33 9. Koynova R, Caffrey M (2002) An index of lipid phase diagrams. Chem Phys Lipids 115:107–219 10. Sari A, Karaipekli A (2012) Fatty acid esters-based composite phase change materials for thermal energy storage in buildings. Appl Therm Eng 37:208–216 11. Irving NM (Rio Oeste S.A.) (2008), CA 2687816 12. Frank D, Metcalfe LD, Park JYG (1986) Process for the separation of amides from nitriles, US 4575434 13. Dubois JL (Arkema) (2010), WO 2010103223 14. Wiley RH, Guierrant WB (1949) Diamides from carboxylic acids and their nitriles. J Am Chem Soc 71(3):981–982
1627 15. Dunn RO (2008) Crystallization behavior of fatty acid methyl esters. J Am Oil Chem Soc 85:961–972 16. Grenet J, Legendre B (2010) Analyse calorime´trique diffe´rentielle a` balayage (DSC). Techniques de l’inge´nieur P1205:1–27 17. Joback KG, Reid RC (1987) Estimation of pure-component properties from group-contributions. Chem Eng Comm 57:233–243 18. Knothe G, Dunn RO (2009) A comprehensive evaluation of the melting points of fatty acids and esters determined by differential scanning calorimetry. J Am Oil Chem Soc 86:843–856 19. Knight JA, Zook HD (1952) Reduction of aliphatic nitriles by the Stephen reaction. J Am Chem Soc 74:4560 20. van Bommel MJ, Oonk HAJ, Cees van Miltenburg J (2004) Heat capacity measurements of 13 methyl esters of n-carboxylic acids from methyl octanoate to methyl eicosanoate between 5 K and 350 K. J Chem Eng Data 49:1036–1042 21. Aleby S, Von Sydow E (1960) The crystal structure of methyl stearate. Acta Cryst 13:487 22. Ishii K, Nakayama H, Koyama K, Yokoyama Y, Ohashi Y (1997) Molecular conformation of butanenitrile in gas, liquid, glass and crystalline states: in relation to the stability of the glass state. Bull Chem Soc Jpn 70:2085–2091 23. Eddleston MD, Lloyd GO, Jones W (2012) Cocrystal dissociation and molecular demixing in the solid state. Chem Commun 48:8075–8077 24. Dreezen G, Groeninckx G, Swier S, Van Mele B (2001) Phase separation in miscible polymer blends as detected by modulated temperature differential scanning calorimetry. Polymer 42:1449–1459 25. Abate L, Badea E, Blanco I, Della Gatta G (2008) Heat capacities and enthalpies of solid–solid transitions and fusion of a series of eleven primary alkylamides by differential scanning calorimetry. J Chem Eng Data 53:959–965
123
