Solubilities of pharmaceutical and bioactive compounds in trihexyl(tetradecyl)phosphonium chloride ionic liquid

Fluid Phase Equilibria 397 (2015) 18–25

Contents lists available at ScienceDirect

Fluid Phase Equilibria j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / fl u i d

Solubilities of pharmaceutical and bioactive compounds in trihexyl (tetradecyl)phosphonium chloride ionic liquid Ricardo A. Faria, Ewa Bogel-Łukasik * REQUIMTE, Departamento de Química, Faculdade de Ciências e Tecnologia, Universidade Nova de Lisboa, 2829-516 Caparica, Portugal

A R T I C L E I N F O

A B S T R A C T

Article history: Received 5 February 2015 Received in revised form 28 March 2015 Accepted 30 March 2015 Available online 2 April 2015

The solubility of pharmaceutical and bioactive compounds, such as N-acetyl-L-cysteine, isoniazid, pyrazine-2-carboxamide, coumarin, 4-hydroxycoumarin, 40 -isobutylacetophenone, ibuprofen and thymoquinone, was tested in trihexyl(tetradecyl)phosphonium chloride. Hydrophobicity/hydrophilicity feature and melting point affected the solubility of the solutes in [P6,6,6,14][Cl]. 40 -Isobutylacetophenone, thymoquinone, coumarin and ibuprofen exhibited the best solubility in the IL due to their hydrophobicity. Then, N-acetyl-L-cysteine was found to be less soluble, and later on isoniazid, 4hydroxycoumarin and pyrazinecarboxamide showed limited solubility in IL. The solid–liquid phase equilibria of all investigated systems were described using the six different correlation equations. Considering the correlation of the phase equilibrium data, the satisfactory results which revealed a good description with an acceptable standard deviation temperature range were collected for systems with: Nacetyl-L-cysteine, coumarin, thymoquinone and ibuprofen. The solubilities of the studied compounds were approximately 2 times higher in trihexyl(tetradecyl)phosphonium chloride than in trihexyl (tetradecyl)phosphonium bis[(trifluoromethyl)sulfonyl]amide ionic liquid. ã 2015 Elsevier B.V. All rights reserved.

Keywords: Solubility Solid–liquid equilibrium Ionic liquid Green solvent Drugs

1. Introduction Volatile organic solvents used in chemical synthesis and pharmaceutical industries generate about 350 million tons of toxic waste per year. Ionic liquids, IL are considered as alternative media that are non-flammable, highly thermally stable [1] and non-volatile green solvents [2]. Ionic liquids properties can be tuned by changing the cation or the anion making them suitable for a specific application, [3–5] such as reaction media, [6–8] active pharmaceutical ingredients [9], catalysts [8] and separation media [10]. ILs have been recently proposed to be used in the pharmaceutical applications [5,11,12]. Sheldon’s E-factor, defined as the mass ratio of waste to desired product, typically reaches E factors of 25– 100 for the pharmaceutical industry, the highest among the oil refining, and the bulk or fine chemicals sectors [13]. For this reason, attention is focused on the development of pharmaceutical processes with waste minimization and assessing its current status in the broad context of green chemistry and sustainability. Particularly, the pharmaceutical industry is seeking for solutions to

* Corresponding author. Tel.: +351 212948500. E-mail address: [email protected] (E. Bogel-Łukasik). http://dx.doi.org/10.1016/j.fluid.2015.03.053 0378-3812/ ã 2015 Elsevier B.V. All rights reserved.

the problem of waste generation in chemicals’ manufacture. Considering the various toxicity of ionic liquids and the commonly used solvents in drug development, the advantageous use of IL comes from their inflammability in a way to be competitive to flammable solvents which use affects the engineering design, features of pharmaceutical facilities and process equipment [5]. It can be desirable to develop manufacturing processes with solvents that are generally identified as being “safer” [14] and suitable in pharmaceutical processing [11]. One of the most important feature of ionic liquids except their non-flammability and non-volatility, is their high solvating power [15] for organic, inorganic and organometallic compounds. Ionic liquids are used in solvent extraction as one of the major hydrometallurgical techniques for the separation and purification of metals [16]. For example, extraction system with trihexyl (tetradecyl)phosphonium chloride was applied in the separation of iron from neodymium and of cobalt from samarium [17] and of cobalt from nickel present in an aqueous phase. The IL’s use was advantageous due to elimination of the volatile and flammable character of the extraction phase. These separations are relevant to recycling of rare earths from permanent magnets [18]. Moreover, trihexyl(tetradecyl)phosphonium chloride was reported to be applied in triphasic catalysis [19], and its use for classic solvent

R.A. Faria, E. Bogel-Łukasik / Fluid Phase Equilibria 397 (2015) 18–25

Nomenclature

[P6,6,6,14] [Cl] [NTf2] DSC SLE T g12–g22,g21–g11 GE K li n q P1, P2 r R SLE Texpi Tcali Tfus,1 Tg u Vm x1 Z

a

g1 DCp,g DfusCp Dg12, Dg21 DfusH1 Dhh Du12, Du21 sT V

Trihexyl(tetradecyl)phosphonium Chloride Bis[(trifluoromethyl)sulfonyl]amide Differential scanning calorimetry Solid–liquid equilibrium Temperature Adjustable parameters of Wilson Gibbs excess energy Association constant Bulk factor Number of experimental points Pure component surface parameter model parameters resulting from the minimization procedure Pure component volume parameter Ideal gas constant Solid–liquid equilibrium Experimental equilibrium temperature Calculated equilibrium temperature Melting temperature of solute Glass transition temperature Standard uncertainty of measurement Molar volume of pure compound at 298.15 K Mole fraction of solute Coordination number Constant of the NRTL, NRTL 1 and NRTL 2 equations Activity coefficient of solute Heat capacity at glass transition temperature Heat capacity between the solid and liquid at the melting temperature Adjustable parameters of NRTL, NRTL 1 and NRTL 2 equations Enthalpy of fusion of solute Enthalpy of association Adjustable parameters of UNIQUAC and UNIQUAC ASM equations Root-mean-square deviation of temperature Objective function

extraction processes is justified from an economical point of view compared to the use of fluorinated ionic liquids [20]. The study on possible recovery of pharmaceutical agents from ionic liquid solvent would be very interesting to perform basing on promising ionic liquid based aqueous biphasic systems as a versatile tool for recovery of antioxidants compounds [21], and of paracetamol from pharmaceutical waste [22]. Supercritical carbon dioxide extraction [23] and pervaporation [24] have been proposed as “green” methods to recover and purify solutes from ILs. There are several pharmaceutical and bioactive agents, that solubilities in ionic liquids were provided, concluding that the ionic liquids are sufficient solvents for the drugs and can be suitable for the pharmaceutical processing [20]. One of them is N-acetyl-L-cysteine (NAC), that is a natural intracellular antioxidant agent important in treatment of HIV infection, cancer, heart diseases, mental illnesses [25]. Two others are antibiotic drugs (isoniazid and pyrazinecarboxamide) [26] used to treat tuberculosis that causes more deaths

19

than AIDS and malaria together. Others are coumarins that found application in pharmaceutical and cosmetic industries. Coumarin based derivatives exhibit anticoagulant, carcinogenic and antibiotic properties. Coumarin can be useful in the antitumour therapy [29], and 4- hydroxycoumarin is an intermediate toward other pharmaceutical compounds, Dicoumarol [27–29]. The other is thymoquinone, an antibiotic and anti-cancerous agent [30–32], and ibuprofen, a non-steroidal anti-inflammatory drug that lowers risk of Parkinson disease. The list closes 40 -isobutylacetophenone that is a precursor in ibuprofen synthesis [33–35]. Solvent screening studies are essential to optimize solvent dependency (solubility, polymorphism, crystallinity and crystal habit of pharmaceutical compounds) [4,11,36]. Solubilities of above stated that solutes have been studied in imidazolium [37–40] and ammonium ILs [41] and phosphonium bis(trifluoromethylsulfonyl) imide [42] proposing them to be alternative solvents to organic toxic solvents routinely used in pharmaceutical industries [35,36,39]. Considering drawback of the price of phosphonium bis (trifluoromethylsulfonyl) imide, and advantageous use of the corresponding chloride ionic liquid as the ionic liquid can easily be generated after extraction in the industrial application [16], we aim to investigate solubilities of eight pharmaceutical or bioactive agents of pharmaceutical and bioactive compounds in chloride IL. The aim of this work is to provide solubility data of isoniazid (IUPAC name: isonicotinohydrazide), pyrazinecarboxamide (IUPAC name: pyrazine-2-carboxamide), N-acetyl-L-cysteine, coumarin, 4hydroxycoumarin, thymoquinone (IUPAC name: 2-isopropyl-5methylbenzo-1,4-quinone), ibuprofen (IUPAC name: 2-(4-isobutylphenyl) propionic acid) and 40 -isobutylacetophenone in trihexyl (tetradecyl)phosphonium chloride, [P6,6,6,14][Cl]. 2. Materials and methods 2.1. Chemicals The chemical structures of solutes and the solvent used in this research are illustrated in Fig. 1. The sample table is depicted in Table 1. 2.2. Methods 2.2.1. Solid–liquid equilibria measurement Solid–liquid equilibria (SLE) of studied systems were obtained at the ambient pressure of 0.1 MPa and at temperature ranging from 275.12 K to 410.26 K using a dynamic (synthetic) method. Experiments were performed in a Pyrex glass cell. The cell could be opened/closed by a Teflon valve at the end of a long, capillary-thin (inner diameter of 0.1 mm) neck. It allowed the cell to be deeply immersed in a temperature-controlled bath, while, at the same time, diminishing losses due to evaporation. The solutions were prepared by weighing the pure components with an accuracy of 104 g. The mixture of solute and solvent was heated very slowly (with maximum heating rate of 2 K h1 near the equilibrium temperature), with continuous stirring. The measurement cell was placed in a thermostatic bath with water (293–333 K) or silicon (333–416 K). The last crystal disappearance temperatures, detected visually, were measured with a calibrated DOSTMANN electronic P600 thermometer equipped in a Pt 100 probe totally immersed in the thermostatic liquid. The uncertainty of the temperature measurements was 0.03 K and that of the mole fraction did not exceed 0.0005. 2.2.2. Differential scanning calorimetry Glass transition temperature, Tg and heat capacity at glass transition temperature, DCp,g, of (trihexyl(tetradecyl)

20

R.A. Faria, E. Bogel-Łukasik / Fluid Phase Equilibria 397 (2015) 18–25

Fig. 1. The structure of the solvent and solutes used in this work. (a) Trihexyl(tetradecyl)phosphonium chloride; (b) isoniazid; (c) pyrazinecarboxamide; (d) coumarin; (e) 4hydroxycoumarin; (f) N-acetyl-L-cysteine; (g) thymoquinone; (h) 40 -isobutylacetophenone; (i) ibuprofen.

phosphonium chloride (Table 2) was acquired using a differential scanning calorimetry (DSC), according to the following procedure: a sample of compound, up to 10 mg, was encapsulated in a 40 mL aluminium crucible. The sample was scanned in a temperature range from 143 to 373 K. All scans were performed at the heating rate of 10 K/min. The results are the average of at least three scans. Prior to each measurement the DSC instrument (Mettler Toledo DSC 822e) was calibrated with a sample of indium and zinc, both with 99.9999 mol% purity. The zinc sample was selected because zinc allows for the calibration in the range of temperatures corresponding to the expected range for high melting point

compounds. The calorimetric accuracy was 1% and the precision was 0.5%. In respect to melting points, the uncertainty was estimated at the level of 0.5 K. 3. Results and discussion 3.1. Solid–liquid equilibrium (SLE) This work is focused on presenting solubility data of pharmaceutical and bioactive compounds in the trihexyl(tetradecyl)phosphonium chloride, [P6,6,6,14][Cl]. The systems studied in

R.A. Faria, E. Bogel-Łukasik / Fluid Phase Equilibria 397 (2015) 18–25

21

Table 1 Characteristics of chemicals used in this study. Chemical name

Source

Initial mass fraction purity

Purification method

Final mass fraction purity

Water content in the final sample/ ppm

Analysis method

[P6,6,6,14][Cl]

Iolitec, Germany

>95 mol%

Degassed, dried, and freed from residues of volatile compounds by 0.1 Pa vacuum at T = 333.15 K for minimum 48 h prior to experiment None None None

99 mol%

150

NMR and coulometric Karl– Fischer titration

99 mol%

500

250 450

N-acetyl-L-cysteine Sigma– Isoniazid Aldrich Pyrazine-2carboxamide (pyrazinecarboxamide) Coumarin 4-Hydroxycoumarin Thymoquinone Ibuprofen TCI Chemicals Alfa Aesar 40 Isobutylacetophenone

99 mol% 99 mol%

99 mol% 98 mol% 99 mol% >98 mol%

None None None None

99 mol% 98 mol% 99 mol% >98 mol%

97 mol%

None

97 mol%

commercial availability [20]. Solubility data obtained here are described by the solid–liquid phase envelopes. The solid–liquid phase diagrams were studied by a dynamic method in a temperature range from 275.12 K to 410.26 K. The thermophysical properties of compounds used in this study are collected in Table 1 of Supporting information. The solubility data of the studied pharmaceutical or bioactive compounds in the aforementioned ionic liquid are presented in

this investigation were composed of the following the pharmaceutical and bioactive compounds, N-acetyl-L-cysteine, isoniazid, pyrazinecarboxamide, coumarin, 4-hydroxycoumarin, 40 -isobutylacetophenone, ibuprofen and thymoquinone, +[P6,6,6,14][Cl]. [P6,6,6,14][Cl] exhibits glass temperature temperate, Tg equaled to 198.88 K and heat capacity at glass transition temperature, DCp,g equaled to 279.44 J mol1 K1. This phosphonium chloride ionic liquid was selected because of its hydrophobic nature and

Table 2 Experimental solubility of N-acetyl-L-cysteine, isoniazid, pyrazinecarboxamide, coumarin, 4-hydroxycoumarin, ibuprofen and thymoquinone (1) in [P6,6,6,14][Cl] (2) at temperature T and at ambient pressure. g1 is the activity coefficient of solute.a [P6,6,6,14][Cl]+ N-acetyl-L-cysteine

Isoniazid

Pyrazinecarboxamide

Coumarin

x1

T/K

g1

x1

T/K

g1

x1

T/K

g1

x1

T/K

g1

0.0034 0.0091 0.0447 0.0887 0.1188 0.1461 0.1768 0.2124 0.3201 0.4545 0.5463 0.6478 0.8419 1.0000

348.78 351.7 363.79 374.10 378.23 378.91 379.09 379.16 379.51 379.75 379.79 379.97 380.04 381.99

112.35 46.02 13.49 9.11 7.61 6.30 5.24 4.37 2.92 2.07 1.73 1.46 1.13 1.00

0.1356 0.1534 0.1678 0.1789 0.2045 0.2844 0.3809 0.5223 1.0000

301.45 333.24 350.67 364.23 378.52 393.51 400.54 409.96 445.84

0.20 0.51 0.77 1.03 1.28 1.29 1.12 0.99 1.00

0.0195 0.0293 0.0499 0.0649 0.0844 0.0912 0.1345 0.1535 1.0000

295.34 312.35 348.81 374.40 394.96 397.45 409.30 410.26 461.42

0.68 0.88 1.71 2.65 3.35 3.28 2.89 2.58 1.00

0.0156 0.0234 0.0567 0.0945 0.1640 0.2343 0.4178 0.6390 0.8923 1.0000

275.12 279.59 296.63 311.95 323.94 327.41 331.96 332.45 333.67 341.68

12.14 9.28 6.21 5.49 4.18 3.16 1.96 1.29 0.95 1.00

[P6,6,6,14][Cl] + Ibuprofen

4-Hydroxycoumarin

Thymoquinone

x1

T/K

g1

x1

T/K

g1

x1

T/K

g1

0.0496 0.0543 0.0760 0.0932 0.1340 0.1964 0.2848 0.3991 0.6545 1.0000

289.95 300.01 326.51 340.92 365.90 389.53 395.56 399.67 401.23 488.30

0.63 0.63 0.63 0.63 0.64 0.65 0.68 0.72 0.84 1.00

0.0827 0.0874 0.1190 0.1698 0.2456 0.3273 0.4390 0.5024 0.5939 0.6589 0.7362 0.8256 0.8599 0.8867 0.9634 1.0000

325.39 326.47 331.38 335.42 337.46 339.46 342.12 342.40 342.95 343.46 343.94 344.23 344.27 344.29 344.34 347.70

6.49 6.33 5.34 4.19 3.07 2.43 1.95 1.72 1.48 1.35 1.22 1.10 1.06 1.03 0.95 1.00

0.0827 0.0874 0.1190 0.1798 0.2456 0.2980 0.3576 0.4278 0.5262 0.6145 0.7274 0.8543 0.9234 0.9754 1.0000

291.39 292.57 295.89 297.42 297.56 297.77 297.80 297.85 297.91 297.92 297.98 298.06 298.97 307.18 316.70

6.64 6.48 5.17 3.56 2.61 2.16 1.81 1.51 1.23 1.05 0.89 0.76 0.72 0.83 1.00

a

Standard uncertainties u are u(T) = 0.03 K, u(x) = 0.0005.

22

R.A. Faria, E. Bogel-Łukasik / Fluid Phase Equilibria 397 (2015) 18–25

Table 2 and Figs. 2–4. Data for 40 -isobutylacetophenone are not inserted in the table as the solubility of 40 -isobutylacetophenone (precursor for the synthesis of ibuprofen) was complete in [P6,6,6,14] Cl in a whole range of the mole fraction of the solute, x1. Direct SLE experimental results, temperatures, T, versus x1 (mole fraction of the solute), for the other investigated systems are collected in Table 2. Solubility of thymoquinone, coumarin and ibuprofen in [P6,6,6,14]Cl is driven by their hydrophobicity in the hydrophobic IL [20]. Hydrophobicity effect is marked on the total solubility of 40 isobutylacetophenone in the hydrophobic IL. Therefore, the solubility is decreasing in the following order: 40 -isobutylacetophenone (complete in all x1) > thymoquinone > coumarin > ibuprofen (see Fig. 2). Verifying solubilities of solutes having a similar melting point, such as pair of coumarin and ibuprofen, the less hydrophobic character of solutes (ibuprofen possessing an additional OH group) affects the trend of solubility

obtained: i.e., at 331 K the solubility of coumarin is x1 = 0.4, while ibuprofen has a 4-fold lower solubility. Solubility of N-acetyl-L-cysteine, isoniazid, 4-hydroxycoumarin and pyrazinecarboxamide (Fig. 3) is decreasing in the following order for temperature below 380 K: isoniazid > 4-hydroxycoumarin > pyrazinecarboxamide and N-acetyl-L-cysteine that can partially be affected by hydrophobicity/hydrophilicity character of the molecules. Theoretically, isoniazid should demonstrate lower solubility compared to 4-hydrocoumarin. Isoniazid and 4-hydroxycoumarin should present the opposite trend detected experimentally, as isoniazid is much more hydrophilic by containing two proton donor groups ready to form hydrogen bonds than 4hydroxycoumarin that has a hydroxyl group. Nevertheless, isoniazid was found to be more soluble than 4-hydroxycoumarin in [P6,6,6,14][Cl]. For temperature range above 380 K two effects are responsible for the solubility behavior: the melting point and hydrophobicity. A predominant effect of the melting point of the solute is visible in the case of the solubility of N-acetyl-L-cysteine that for the temperature above 380 K has the best solubility in [P6,6,6,14][Cl] due to the lowest melting point, in spite of being more hydrophilic than isoniazid. The solubility of isoniazid in IL should be superior to N-acetyl-L-cysteine as N-acetyl-L-cysteine has more proton donating groups thus is less hydrophobic than isoniazid, nevertheless such a difference in solubility trend is caused by the difference in the melting points of the solutes. The lower melting point of isoniazid (445.84 K) compared to 4-hydrocoumarin (488.30 K) directs the trend of solubility in the investigated hydrophobic ionic liquid, keeping the same order of the enthalpies of fusion (DfusH4HC = 24.49 kJ mol1 and DfusHISO = 27.91 kJ mol1). It can be noticed that the solubility data are more plateau for 4hydrocoumarin than for isoniazid in the considered temperature range. On the other hand, it is interesting to emphasize that pyrazinecarboxamide, which is less hydrophilic than 4-hydroxycoumarin, is in fact less soluble in [P6,6,6,14][Cl] than 4-hydroxycoumarin, albeit it has significantly lower melting point than 4HC (461.42 K and 488.30 K, respectively). The solubility of all solutes increases slightly with an increase of temperature when the mole fraction is small, while for high solute mole fraction, the insignificant change of temperature affects the solubility more significantly. It is especially visible for the zone

Fig. 3. Comparison of solubilities of N-acetyl-L-cysteine (^, dashed-dotted), isoniazid (5, solid), 4-hydroxycoumarin (&, dotted) and pyrazinecarboxamide (!, short dashed) in [P6,6,6,14][Cl]. Lines are given only as a guide to an eye.

Fig. 4. Comparison of solubilities of 4-hydroxycoumarin in [P6,6,6,14][Cl] (&, dotted) and [P6,6,6,14][NTf2] (&, dashed) and of ibuprofen in [P6,6,6,14][Cl] (*, short dashed) and [P6,6,6,14][NTf2] (, dashed-dotted). Lines are given only as a guide to an eye.

Fig. 2. Comparison of solubilities of thymoquinone (~, dashed), coumarin (&, dashed-dotted-dotted), ibuprofen (*, short dashed) in [P6,6,6,14][Cl]. Lines are given only as a guide to an eye.

R.A. Faria, E. Bogel-Łukasik / Fluid Phase Equilibria 397 (2015) 18–25

23

Table 3 Comparison of solubilities of N-acetyl-L-cysteine, isoniazid, pyrazinecarboxamide, coumarin, 4-hydroxycoumarin, ibuprofen and thymoquinone (1) in [P6,6,6,14][Cl] (2) or [P6,6,6,14][NTf2] [40] at selected temperatures T and at ambient pressure. N-acetyl-L-cysteine T/K

Isoniazid

T/K x1 in [P6,6,6,14][Cl] x1 in [P6,6,6,14] [NTf2] [40]

353.15 0.0136 363.15 0.0426 373.15 0.0830

0.0064 0.0212 0.0417

Pyrazinecarboxamide

303.15 0.1363 313.15 0.1417 323.15 0.1476

0.0678 0.0729 0.0783

Coumarin

T/K x1 in [P6,6,6,14][Cl] x1 in [P6,6,6,14] [NTf2] [40]

x1 in [P6,6,6,14][Cl] x1 in [P6,6,6,14] T/K [NTf2] [40]

298.15 0.0211 303.15 0.0240 313.15 0.0298

0.0130 0.0136 0.0149

Ibuprofen

4-Hydroxycoumarin

x1 in [P6,6,6,14][Cl] x1 in [P6,6,6,14] [NTf2] [40]

298.15 0.0591 303.15 0.0708 313.15 0.0995

0.0301 0.0370 0.0112

Thymoquinone

T/K

x1 in [P6,6,6,14][Cl]

x1 in [P6,6,6,14][NTf2] [40]

T/K

x1 in [P6,6,6,14][Cl]

x1 in [P6,6,6,14][NTf2] [40]

T/K

x1 in [P6,6,6,14][Cl]

x1 in [P6,6,6,14][NTf2] [40]

303.15 313.15 323.15

0.0571 0.0641 0.0723

0.0313 0.0358 0.0406

333.15 338.15 343.15

0.1347 0.2704 0.6190

0.0782 0.1487 0.3256

298.15 303.15 313.15

0.8637 0.9599 0.9916

0.0870 0.9247 0.9758

where liquid–liquid phase equilibria should be expected; however this was not observed experimentally. All examined systems are described by the classical solid–liquid phase envelopes with the eutectic point below the detection limit. As the eutectic point is strongly shifted toward low values of the solute mole fraction, due to the high melting points and enthalpies of fusion of the studied compounds, was not detected experimentally. Concluding solubility dependence obtained in this work the two major factors can be responsible for the solubility of the examined solutes in the IL: hydrophobicity and melting points. Among all solubilities provided in this work, thymoquinone (Fig. 2) was found to exhibit the highest solubility in trihexyl (tetradecyl)phosphonium chloride. This can be caused by the lowest melting point among all solutes and by highly hydrophobic attitude of thymoquinone in a hydrophobic chloride phosphonium ionic liquid. Comparison of solubilities of very structurally similar compounds, such as coumarin (Table 4 and Fig. 2) and 4-hydroxycoumarin (Table 2 and Fig. 3) that are planar and rigid with an aromatic ring, revealing a difference that the OH group confers 4hydroxycoumarin higher hydrophilicity and displays lower solubility in the hydrophobic [P6,6,6,14][Cl] ionic liquid. Taking into consideration the solubilities obtained in this work and reported for [P6,6,6,14][NTf2] [40], the change for chloride anion allowed to obtain better solubilities for all solutes. The comparison of the effect of the anion of trihexyl(tetradecyl)phosphonium IL is presented on example of the solubility of 4-hydroxycoumarin and ibuprofen in both ILs in Fig. 4. Satisfactory results on solubilities obtained in this work seems to be perspective in use of chloride phosphonium IL due to an economical point of view [20]. Table 3 present the comparison of the solubility data (calculated on a base of experimentally determined data here and in [36]) dependent on the anion of ILs studied for the selected temperatures. The solubility of N-acetyl-L-cysteine in [P6,6,6,14][Cl] at the chosen temperatures had a 2-fold increase compared to the solubility in [P6,6,6,14][NTf2]. This increase in solubility was

observed for all the compounds with the exception of thymoquinone. Thymoquinone showed a 10-fold increased solubility in [P6,6,6,14][Cl] at 298.15 K, while for 303.15 and 313.15 K the difference in solubility was very small. This high increase can be explained by the fact the solubility reaches a plateau at lower temperatures in [P6,6,6,14][Cl], meaning in [P6,6,6,14][Cl] the plateau is reached at approximately 297 K, in [P6,6,6,14][NTf2] it is reached at approximately 299 K. 3.2. Correlation of (solid + liquid) equilibrium The ideal solubility of a solid in a liquid can be calculated admitting that total miscibility (x = 1) is obtained at the normal melting temperature of the solid, and that the chemical potential of the solute in the liquid solution at any lower temperature obeys the expression for an ideal solution. For a non-ideal solution, the solute concentration (mole fraction) is replaced by the activity and the solubility is given by the following equation:
 D H1 1 1  ln x1 ¼ fus T T R  1
fus;1   T fus;1 Dfus C p1 T1 ln  1 þ ln g 1; (1) þ  R T fus;1 T1 where x1, g1, DfusH1, DfusCp1, Tfus,1 and T1 stand for mole fraction, activity coefficient, enthalpy of fusion, difference in solute heat capacity between the solid and liquid at the melting temperature, melting temperature of the solute and equilibrium temperature, respectively. The second term of the right side of the equation is a usually small corrective term that accounts for the change of the enthalpy of fusion with temperature. As such data on DfusCp1 are not available for solutes studied in this work, the equation was used in the calculations without the second term. The omission of this term in the practical application of correlation of the SLE data leads to less accurate description of the obtained results, therefore, the standard deviations of the correlation equations are higher

Table 4 Values of enthalpy of association Dhh (kJ mol1) at 323.15 K for the corresponding compounds and the optimised association constants for each solute–solvent pair. [P6,6,6,14][Cl] +

Dhh (kJ mol1) K a

N-acetyl-L-cysteine [20,40] K

Isoniazid [19,23,40] K

Pyrazinecarboxamide [3,23,40] K

Coumarin [20,40] K

4-Hydroxycoumarin [20,40] K

Ibuprofen [40] K

Thymoquinone [40] K

21.00

9.29313.15 K

9.29313.15 K

21.00

21.00

21.00

21.00

5.67

5.52

22.13

Optimized in this work.

12.50

12.50

6.19

a

38.05

24

R.A. Faria, E. Bogel-Łukasik / Fluid Phase Equilibria 397 (2015) 18–25

than for systems with the DfusCp1 included. Nevertheless, due to the limited data, this approach was already successfully used in the past and the correlation gave acceptable results [15–17]. The aforementioned equation may be used assuming the simple eutectic mixtures with a full miscibility in the liquid and immiscibility in the solid phases. In this study, six methods were used to derive the solute activity coefficients g 1 from the so-called correlation equations that describe the Gibbs excess energy (GE), the Wilson [23], UNIQUAC [24], UNIQUAC ASM [25], NRTL [26], NRTL 1 [26] and NRTL 2 [26]. The exact mathematical forms of the equations were presented elsewhere [43]. The two adjustable parameters of the equations were found by an optimisation technique using Marquardt’s maximum likelihood method of minimisation described by the equation presented in the following manner: i2 h n V ¼ Si¼1 T exp  T cal (2) i ðx1 ; P 1 ; P 2 Þ i where V is the objective function, n is the number of experimental points, T exp and T cal denotes respectively experimental and i i calculated equilibrium temperature corresponding to the concentration. The x1, P1 and P2 are model parameters resulting from the minimisation procedure. The root-mean-square deviation of temperature was defined as follows: 2

 2 31=2 exp cal 6 n Ti  Ti 7 s T ¼ 4Si¼1 5 n2

(3)

the pure component parameters r (volume parameter) and q (surface parameter) in the UNIQUAC ASM, NRTL, NRTL 1 and NRTL 2 equations were obtained by means of the following simple

relationships [28]: ri ¼ 0:029281V m

qi ¼

(4)

ðZ  2Þri 2ð1  li Þ þ Z Z

(5)

where Vm is the molar volume of pure compound i at 298.15 K. The molar volume of solute Vm1 (298.15 K) calculated by the group contribution method [29] were used: 156.3 cm3 mol1 for coumarin [36,40], 112.13 cm3 mol1 for 4-hydroxycoumarin [36,40], 126.11 cm3 mol1 for N-acetyl-L-cysteine [36,40], 190.80 cm3 mol1 for ibuprofen [40], 142.78 cm3 mol1 for thymoquinone [40], 94.5 cm3 mol1 for pyrazinecarboxamide [5,7,40] and 110.10 cm3 mol1 for isoniazid [5,7,40]. Values of enthalpy of association were taken from literature (isoniazid, pyrazinecarboxamide [5,8,40], Nacetyl-L-cysteine, coumarin, 4-hydroxycoumarin [38,40], ibuprofen and thymoquinone [40]). For all the solutes, the calculations of the SLE model parameters were carried out using the association constants from literature [36]. For 4-hydroxycoumarin with the reported K the correlations did not gave acceptable standard deviations. Therefore, K had to be optimized in this study and equalled to 38.05. The values of enthalpies of association and association constants are provided in the Table 4. Furthermore, the molar volume of the solvent used for the correlations was 583.56 cm3 mol1 [44]. The Kretschmer–Wiebe model of association for the developing of two adjustable parameters was used [32]. In this work, the value of parameter a, a proportionality constant similar to the nonrandomness constant of the UNIQUAC ASM, NRTL, NRTL 1 and NRTL 2 equations was used in calculations for different binary systems. Values of model parameters obtained by fitting to the solubility results are presented in Table 5 with the corresponding standard deviations for the investigated systems of a hydrophobic chloride ionic liquid and N-acetyl-L-cysteine, or

Table 5 Correlation of the solubility data, SLE, of studied solute (1) + [P6,6,6,14][Cl] (2) by Wilson, UNIQUAC, UNIQUAC ASM, NRTL, NRTL 1, and NRTL 2 equations: values of parameters and deviations. Solute

N-acetyl-L-cysteine Isoniazid Pyrazine carboxamide Coumarin 4-Hydroxycoumarin Ibuprofen Thymoquinone

a b c d e f g h i j

Parameters

Deviations

Wilson

UNIQUAC

g12–g22 g21–g11 J mol1

Du12 Du21

15007.14 2516.36 740.35 88975.98 –

2099.46 6917.30 10937.89 2501.52 130691.78 805.45 1458.35 3830.96 3164.86 6790.89 1273.33 3148.49 1811.81 4910.94

6650.41 1608.14 911.57 829.76 7150.65 2376.36 65549.55 1874.85

J mol1

UNIQUAC ASM Du12 Du21 J mol1 2371.70 5552.00 – 103701.25 2869.48 874.78 1572.40 – 839.79 1629.90 –

NRTL

NRTL 1

NRTL 2

Wilson

UNIQUAC

Dg12 Dg21

Dg12 Dg21

Dg12 Dg21

s Ta

s Ta

J mol1

K

838.17 6645.93 12649.07 1848.62 35046.19 2659.84 12911.86 9453.68 99651.92 6419.13 2409.65 1097.93 450.64 552.71

4279.42 7496.88 8484.33 7881.81 117601.26 2708.74 9916.96 9902.59 9454.83 160.79 2429.82 2466.83 1946.74 1885.33

1011.13 3268.26 7017.65 9095.87 22326.31 7403.48 8656.94 10812.92 –

4.08

281.62 951.40 1039.51 2248.95

J mol1

 2 1=2 n ðT exp T cal Þ According to the equation s T ¼ Si¼1 i n2i . Calculated with the third non randomness parameter a = 0.99. Calculated with the third non randomness parameter a = 0.70. Calculated with the third non randomness parameter a = 0.15. Calculated with the third non randomness parameter a = 0.65. Calculated with the third non randomness parameter a = 0.05. Calculated with the third non randomness parameter a = 0.25. Calculated with the third non randomness parameter a = 0.30. Calculated with the third non randomness parameter a = 0.75. Calculated with the third non randomness parameter a = 0.45.

J mol1

UNIQUAC ASM

NRTL

NRTL 1

NRTL 2

s Ta

s Ta

s Ta

s Ta

K

K

K

K

K

11.93

10.08

4.14b

6.31b

9.72b

20.67

18.58

–

14.35b

12.54c

22.05d

–

17.62

20.80

7.19e

21.17f

9.86d

2.82

2.97

3.67

2.74f

2.16g

2.43f

21.43

17.63

–

24.09g

22.70h

–

0.78

1.78

2.28

2.03b

2.00i

2.29b

6.70

5.43

–

7.17b

7.06j

3.58b

R.A. Faria, E. Bogel-Łukasik / Fluid Phase Equilibria 397 (2015) 18–25

isoniazid, or pyrazinecarboxamide, or coumarin, or 4-hydroxycoumarin, or ibuprofen, or thymoquinone. The data presented in Table 2 shows the solubility results as well as the activity coefficients for systems of N-acetyl-L-cysteine, or isoniazid, or pyrazinecarboxamide, or coumarin, or 4-hydroxycoumarin, or ibuprofen, or thymoquinone and [P6,6,6,14][Cl]. Analysis of the obtained results allows to conclude that the solubility is clearly lower than the ideal solubility for N-acetyl-Lcysteine, coumarin, ibuprofen due to hydrophobic relation. It is visible for thymoquinone for x1 < 0.7, where later melting point effect is influencing. This confirms that obtained solubility data of the hydrophobic solutes in hydrophilic ionic liquid and of the hydrophilic solutes in hydrophobic ionic liquids are lower than solubility of the pair hydrophilic solute–solvent (e.g., 4-hydroxycoumarin + [C2mim][OTf]) or the hydrophobic solute–solvent (coumarin + [C10mim][NTf2]) as reported in the literature [36]. Considering that the systems investigated in this work are complex and the assumption related to the association constants of the examined pharmaceutical and bioactive compounds, we can conclude that the correlation equations gave acceptable s T. Therefore, the satisfactory results which revealed a good description with an acceptable standard deviation temperature range were collected for systems with: N-acetyl-L-cysteine, coumarin, thymoquinone and ibuprofen. The best standard deviation for systems with N-acetyl-L-cysteine (4.08 K) and ibuprofen (0.78) was obtained using Wilson equation. NRTL1 equation gave the lowest s T, for system with coumarin (2.16 K), and NRTL2 allowed to obtained s T equalled to 3.58 K with thymoquinone. The average standard deviation (s T), obtained in the correlation of the experimental results ranges from 0.78 to 24.09 K, but s T above 7 are not taken into consideration for discussion of the results, but are compiled in the Table 4. For systems of isoniazid, pyrazinecarboxamide and 4-hydroxycoumarin, the s T was higher than 7 K for all correlation equations used. 4. Conclusions Solubility data for several pharmaceutical and bioactive compounds was provided in a hydrophobic trihexyl(tetradecyl) phosphonium chloride ionic liquid. Solubility of solutes increased slightly with an increase of temperature for low mole fraction of solute and for plateau zone the increase of solubility is significant for even minimal increase of temperature. The solubility of most of the compounds was in satisfactory ranges. Pyrazinecarboxamide has the lowest solubility, and thymoquinone exhibited the best solubility, while 40 -isobutylacetophenone was found to be completely miscible in the IL. The most soluble pharmaceutical and bioactive compounds were found for [P6,6,6,14][Cl]. The solubilities of all drugs in trihexyl(tetradecyl)phosphonium chloride were approximately 2 times higher than in the trihexyl (tetradecyl)phosphonium bis[(trifluoromethyl)sulfonyl]amide ionic liquid. Acknowledgement This work was financed by national funding from Fundação para a Ciência e Tecnologia – FCT (Portugal), Universidade Nova de Lisboa, through project PEst-C/EQB/LA0006/2013 and IF Investigator FCT 2013 IF/01643/2013. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.fluid.2015.03.053.

25

References [1] T. Fukushima, A. Kosaka, Y. Ishimura, T. Yamamoto, T. Takigawa, N. Ishii, T. Aida, Science 300 (2003) 2014–2072. [2] D.D. Patel, J. Lee, Chem. Rec. 12 (2012) 329–355. [3] E.I. Alevizou, E.C. Voutsas, J. Chem. Thermodyn. 62 (2013) 69–78. [4] C.M.S.S. Neves, M.L.S. Batista, A.F.M. Claudio, L.M.N.B.F. Santos, I.M. Marrucho, M.G. Freire, J.A.P. Coutinho, J. Eng. Chem. Data 55 (2010) 5065–5073. [5] C. Lourenço, C.I. Melo, R. Bogel-Łukasik, E. Bogel-Łukasik, J. Eng. Chem. Data 57 (2012) 1525–1533. [6] V.I. Parvulescu, C. Hardacre, Chem. Rev. 107 (2007) 2615–2665. [7] C.I. Melo, R. Bogel-Łukasik, E. Bogel-Łukasik, J. Supercrit. Fluid 61 (2012) 191–198. [8] E. Bogel-Łukasik, S. Santos, R. Bogel-Łukasik, M. Nunes da Ponte, J. Supercrit. Fluid 54 (2010) 210–217. [9] W.L. Hough, M. Smiglak, H. Rodriguez, R.P. Swatloski, S.K. Spear, D.T. Daly, J. Pernak, J.E. Grisel, R.D. Carliss, M.D. Soutullo, J.H. Davis, R.D. Rogers, New J. Chem. 31 (2007) 1429–1436. [10] A. Arce, M.J. Earle, S.P. Katdare, H. Rodríguez, K.R. Seddon, Fluid Phase Equilibr. 261 (2007) 427–433. [11] K.B. Smith, R.H. Bridson, G.A. Leeke, J. Chem. Eng. Data 56 (2011) 2039–2043. [12] H. Mizuuchi, V. Jaitely, S. Murdan, A.T. Florence, Eur. J. Pharm. Sci. 33 (2008) 326–331. [13] R.A. Sheldon, Chem. Ind.-London (1992) 903–906. [14] S. Pareek, C. Rajsharad, R. Kirkinde, Safer alternatives to organic solvents, Exp. Pharma (2008) . [15] R. Bogel-Łukasik, D. Matkowska, M.E. Zakrzewska, E. Bogel-Łukasik, T. Hofman, Fluid Phase Equilibr. 295 (2010) 177–185. [16] S. Wellens, R. Goovaerts, C. Mooller, J. Luyten, B. Thijs, K. Binnemans, Green Chem. 15 (2013) 3160–3164. [17] T. Vander Hoogerstraete, S. Wellens, K. Verachtert, K. Binnemans, Green Chem. 15 (2013) 919–927. [18] K. Binnemans, P.T. Jones, B. Blanpain, T. Van Gerven, Y. Yang, A. Walton, M. Buchert, J. Clean. Prod. 51 (2013) 1–22. [19] K. Anderson, H. Rodriguez, K.R. Seddon, Green Chem. 11 (2009) 780–784. [20] S. Wellens, B. Thijs, K. Binnemans, Green Chem. 14 (2012) 1657–1665. [21] J.H. Santos, F.A. e Silva, S.P.M. Ventura, J.A.P. Coutinho, R.L. de Souza, C.M.F. Soares, A.S. Lima, Biotechnol. Progr. 31 (2015) 70–77. [22] F.A. e Silva, T. Sintra, S.P.M. Ventura, J.A.P. Coutinho, Sep. Purif. Technol. 122 (2014) 315–322. [23] L.A. Blanchard, J.F. Brennecke, Ind. Eng. Chem. Res. 40 (2001) 287–292. [24] T. Schäfer, C.M. Rodrigues, C.A.M. Afonso, J.G. Crespo, Chem. Commun. 17 (2001) 1622–1623. [25] M. Berk, D.L. Copolov, O. Dean, K. Lu, S. Jeavons, I. Schapkaitz, M. AndersonHunt, A.I. Bush, Biol. Psychiatry 64 (2008) 361–368. [26] M. Dolezal, M. Miletin, J. Kunes, K. Kralova, Molecules 7 (2002) 363–373. [27] M.A.R. Matos, C.C.S. Sousa, M.S. Miranda, V.M.F. Morais, J.F. Liebman, J. Phys. Chem. B 113 (2009) 11216–11221. [28] C.C.S. Sousa, V.M.F. Morais, M.A.R. Matos, J. Chem. Thermodyn. 42 (2010) 1372–1378. [29] A. Maucher, M. Kager, E. Angerer, J. Cancer Res. Clin. Oncol. 119 (1993) 150–154. [30] S. Hossain, E. Sikes-Thurston, S.H. Leppla, A.N. Wein, The Journal of Experimental Secondary Science, copyright 2011–2012, ISSN# 2162-8092 [31] A.M. Abdel-Fattah, K. Matsumoto, H. Watanabe, Eur. J. Pharm. Sci. 400 (2000) 89–97. [32] H. Gali-Muhtasib, A. Roessner, R. Schneider-Stock, Int. J. Biochem. Cell Biol. 38 (2006) 1249–1253. [33] A.I. Gasco-Lopez, R. Izquierdo-Hornillos, A. Jimenez, J. Pharm. Biomed. 21 (1999) 143–149. [34] M. Cleij, A. Archelas, R. Furstoss, J. Org. Chem. 64 (1999) 5029–5035. [35] R.A. Kjonass, P.E. Williams, D.A. Counce, L.R. Crawley, J. Chem. Educ. 88 (2011) 825–828. [36] T. Lee, C.S. Kuo, Y.H. Chen, Solubility, Pharm. Technol. 30 (2006) 72–92. [37] A. Forte, C. Melo, I.R. Bogel-Łukasik, E. Bogel-Łukasik, Fluid Phase Equilibr. 318 (2012) 89–95. [38] A.D. Santos, A.R.C. Morais, C. Melo, R. Bogel-Łukasik, E. Bogel-Łukasik, Fluid Phase Equilibr. 356 (2013) 18–29. [39] A.I. Gasco-Lopez, R. Izquierdo-Hornillos, A. Jimenez, Fluid Phase Equilibr. 318 (2012) 89–95. [40] S. Hossain, E. Sikes-Thurston, S.H. Leppla, A.N. Wein, The Journal of Experimental Secondary Science, copyright 2011–2012, ISSN# 2162-8092. [41] C.I. Melo, R. Bogel-Łukasik, M. Nunes da Ponte, E. Bogel-Łukasik, Fluid Phase Equilibr. 338 (2013) 209–216. [42] R.A. Faria, M. Nunes da Ponte, E. Bogel-Łukasik, Fluid Phase Equilibr. 385 (2015) 1–9. [43] U. Domanska, J. Rolinska, Fluid Phase Equilibr. 86 (1993) 233–250. [44] C.M.S.S. Neves, P.J. Carvalho, M.G. Freire, J.A.P. Coutinho, J. Chem. Thermodyn. 43 (2011) 948–957.