Deepening of the Role of Cation Substituents on the Extractive Ability

Publication Date (Web): January 12, 2017. Copyright © 2017 American Chemical Society. *Phone: +34 986812290. E-mail: [email protected]. Cite this:ACS ...
0 downloads 0 Views 3MB Size
Research Article pubs.acs.org/journal/ascecg

Deepening of the Role of Cation Substituents on the Extractive Ability of Pyridinium Ionic Liquids of N‑Compounds from Fuels Pedro Verdía,§ Emilio J. González,† Daniel Moreno,‡ José Palomar,‡ and Emilia Tojo*,§

Downloaded via NEW MEXICO STATE UNIV on July 4, 2018 at 18:55:39 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

§

Organic Chemistry Department, Faculty of Chemistry, University of Vigo, Campus Lagoas Marcosende, 36210 Vigo, Pontevedra, Spain † Departamento de Ingeniería Química Industrial y del Medio Ambiente, Universidad Politécnica de Madrid, Calle de José Gutiérrez Abascal 2, 28006 Madrid, Spain ‡ Departamento de Química Física aplicada, Universidad Autónoma de Madrid, Campus de Cantoblanco, C/Francisco Tomás y Valiente 7, 28049 Madrid, Spain S Supporting Information *

ABSTRACT: New functionalized 1-butyl-3-Xpyridinium dicyanamide ionic liquids (ILs) were synthesized by adding groups (X = cyano, amino, chlorine, alkyl) with different substituent effects. Experimental and computational analyses were performed to evaluate the role of the pyridinium substituent on IL solvent properties, particularly as an extracting agent of pyridine from fuels. Quantum chemical calculations and NMR measurements indicated that the hydrogen bond (HB) donor character of the cation was successfully tuned by an adequate substitution. The COSMO-RS study showed that pyridine produces exothermic mixtures with these ILs, mainly due to favorable HB interactions. However, the mixture behavior was found to be controlled by entropy. Experimental and calculated Liquid−Liquid Equilibrium (LLE) data of pyridineheptane-IL mixtures revealed that the new functionalized ILs present favorable partition coefficients and selectivity for extracting N-compounds from aliphatic mixtures. It was possible to enhance the solvent performance by using tetraalkyl substituents, which increases the entropy of the system. KEYWORDS: Ionic liquids, Functional groups, Pyridine liquid−liquid extraction, COSMO-RS, Solvent properties



fuels.8,9 However, the greatest difficulty is to find an adequate extractive solvent, very selective for the extraction of Ncompounds (without affecting the olefin content) and easy to be recovered after the extraction step. The possibility of using ionic liquids (ILs) as solvents for “clean” liquid−liquid extraction was proposed in 1998 by Huddleston et al.,10 who found that these liquid salts may be a suitable media for the design of novel liquid−liquid extraction systems. Their unique properties such as a negligible vapor pressure, good thermal stability, wide liquid range, ability to dissolve a wide range of materials, and the possibility of designing their structures for each specific application make them an excellent choice to get more efficient and environmentally friendly separation processes.11,12 Since then, the use of ILs as solvents in liquid extraction has been applied to separate different aromatic compounds, including nitrogen compounds.13−23 Recent Aspen Plus supported conceptual design of aromatic−aliphatic separation from naphtha indicated favorable performance of ILs as extracting solvents. This allows obtaining aliphatic and aromatic products with high purity and

INTRODUCTION The presence of nitrogen compounds in diesel and gasoline leads to emission of greenhouse gases (NOx), which contributes to acid rain, global warning, and destruction of the ozone layer. In addition, NOx is linked with a number of adverse effects on the human respiratory system. As a consequence, current legislation has set very strict limits to the content of nitrogen compounds in fuels.1,2 The nitrogen compounds present in fuels are mainly amines, nitriles, and heterocyclic aromatic compounds, as pyrrole or pyridine.3,4 At an industrial level, the conventional way for reducing this N-content is hydrodenitrogenation (HDN), a reduction reaction at high pressure and temperature (300−450 °C) in the presence of hydrogen and catalysts to produce NH3.5,6 This process is extremely expensive, and although it works for aliphatic N compounds, it is inefficient in the reduction of aromatic compounds, that even in small amounts may saturate the catalyst surface decreasing the HDN process effectiveness.7 For all these reasons, the development of new approaches to reduce the N content in fuel oils is needed. Since liquid−liquid extraction operates at mild conditions and does not change the chemical constituents of the species involved, it is a promising method that can be used for desulfurization and denitrogenation of gasoline and diesel © 2017 American Chemical Society

Received: December 1, 2016 Revised: December 30, 2016 Published: January 12, 2017 2015

DOI: 10.1021/acssuschemeng.6b02922 ACS Sustainable Chem. Eng. 2017, 5, 2015−2025

Research Article

ACS Sustainable Chemistry & Engineering

ring and pyridine ring.34,36 Although anions or cations based on the −CN functional group were not included in the reported theoretical analysis, it has been experimentally stated that it enhances the extraction properties of ILs. In addition, whereas the above computational studies were focused on the interaction energies, the entropic contribution to the mixture behavior was not considered. However, recent experimental evidence indicated that they may be relevant in IL-Nheterocyclic compound mixtures.36 With the aim of providing a better understanding of the interaction mechanisms between ILs and N-compounds in fuels, in this work a series of ILs derived from 1butylpyridinium cations substituted with different electron donor and withdrawing groups were tested as potential solvents to extract N-heterocyclic compounds from fuel oils. Heptane and pyridine were selected as representative of aliphatic and basic N-compounds, respectively. Taking into account the good results previously obtained with the [DCA] anion,31 this one was selected. Four 1-butylpyridinium based ILs with different functional groups (−CN, −Cl, −NH2, −CH3) on position 3 were synthesized: 1-butyl-3-cyanopyridinium dicyanamide [1B3CNPy][DCA], 1-butyl-3-chloropyridium dicyanamide [ 1 B 3 ClPy][DCA], 3-amino-1-butylpyridium dicyanamide [1B3NH2Py][DCA], and 1-butyl-3-methylpyridinium dicyanamide [1B3MPy][DCA]. The IL 1-butyl-3,5-dimethyl-2-pentylpyridinium dicyanamide [1B3M5M2PPy][DCA], previously studied by our group,31 has also been included in this work for comparison purposes. A detailed computational analysis based on density functional theory was performed to evaluate the effect of the added functional groups on molecular and solvent properties of these ILs. In addition, the thermodynamics of the corresponding pyridine-IL and heptane-IL binary mixtures was studied by means of the quantum chemicalfounded COSMO-RS method, obtaining deeper insight into the interaction energy, enthalpy, entropy, and free energy of these systems. Then, the liquid−liquid equilibrium data for heptane + pyridine + ionic liquid ternary mixtures were experimentally and theoretically determined at T = 298 K and at atmospheric pressure. The extractive denitrogenation ability of the new 1-butylpyridinium based ILs with different functional groups was analyzed in terms of the estimated values of partition coefficient and selectivity.

low energy expenses, which is mainly related to the very low vapor pressure of ILs and their easy regeneration by flash distillation.24,25 Regarding the separation of nitrogen compounds, some of the most recent and significant contributions are the following. Anantharaj and Banerjee9 carried out a computational screening of 168 potential ILs for the removal of aromatic nitrogen, concluding that the presence of N and O into the cations combined with anions incorporating S and F atoms greatly improves the hydrogen bonding with the aromatic nitrogen compound, thereby increasing selectivity. In 2013, the studies developed by Gabrić et al.8 indicated that the pyridinium-based IL (1-hexyl-3,5-dimethylpyridinium bis(trifluoromethylsulfonyl)imide, [C6MMPy][Tf2N]) was an effective selective solvent for denitrogenation of model diesel. Chen et al.26 demonstrated that some acidic ionic liquids are capable of extracting cyclic N-compounds, and Hizaddin et al.27 reported that some sulfate-based ILs are also highly effective solvents for use in extractive denitrogenation of diesel for both 5-membered and 6-membered nitrogen compounds. More recently, other ILs containing imidazolium-/morpholinium-/ pyridinium-/quinolinium-/isoquinolinium cations were proposed as solvents to extract pyridine and its derivatives from fuels oils.28−32 The influence of anion nature on pyridine separation capacity and selectivity has been also evaluated.18,31 This available experimental evidence indicates that the selection of both cation and anion plays a main role in the IL solvent properties for extracting cyclic N-compounds from aliphatic hydrocarbons. Aromatic and protic cations, as imidazolium and pyridinium, are favorable selections to increase the separation capacity, while adding alkyl or benzyl substituents to the cation structure enhances the selectivity for pyridine.18,31,32 Furthermore, the use of small and polar CN-functionalized anions, as dicyanamide [DCA] or tricyanomethanide [TCM], significantly improves the N-compound extracting properties of ILs.18,29,31 Recent works by Domanska et al.29,33 revealed that the presence of the −CN group at the pyridinium cation confers promising properties to ILs for the separation of sulfur or nitrogen compounds, enhancing the transport properties of the solvent and the solute−solvent interaction. Abro et al.21 has recently published a review on the extractive denitrogenation of fuel oils using ILs. In order to explain the interaction between ILs and nitrogen compounds, computational studies via quantum chemical methods were also conducted.34−37 Density functional calculations revealed that the predominant cation−anion interaction was not destroyed by the presence of pyridine, indicating the prevalence of ion-pair structures in the ILpyridine solution.34,35 Hydrogen bonding denotes crucial interactions between the IL and pyridine, which are not present in IL-aliphatic hydrocarbon mixtures, explaining the high denitrogenation capacity of ILs.34,35 The COSMO-RS method38 was successfully applied to predict liquid−liquid equilibrium data of IL/N-compound/aliphatic hydrocarbon ternary systems, screening over a wide variety of cations and anions to select the best ILs for extracting N-heterocyclic compounds.9,36,37,39 The general recommendation was to select ILs with an aromatic cation, combined with small and polar anions (as ethylsulfate or acetate) to increase the extraction capability with heterocyclic nitrogen compounds at 298.15 K.9,36,37,39 However, the molecular interaction mechanisms involved in these extractions are not yet fully understood and more research is needed. There is controversy over the occurrence of π−π or CH-π interactions between the cation



MATERIALS AND METHODS

Chemicals. Acetonitrile (Sigma-Aldrich, ACS reagent, ≥99%), acetone (Fischer Chemical, laboratory reagent grade, −Cl > −CH3 > −NH2. Quantum chemical calculations, performed by the GIAO method at the B3LYP/6-311++G** computational level, show good agreement with the experimental trend. These results indicate stronger hydrogen bond interaction between a cation and anion when the pyridinium ring presents electron acceptor substituents with high inductive field effects (as −CN or −Cl, see Taft’s substituents parameters in Table S1 of the Supporting Information). This is ascribed to the higher acidic character of pyridinium H-2 when −CN or −Cl are located in ortho. In contrast, electron donor groups, such as −NH2, decrease the cation−anion hydrogen bond strength with respect to the benchmark case with only a butyl substituent. It should be noted, however, that the hydrogen atoms of −NH2 may form strong hydrogen bonds with the anion, attending to the higher 1H NMR chemical shifts in Table 1. The substitution with additional alkyl groups in the pyridinium ring (see [1B3M5M2PPy][DCA] in Table 1) does not imply strong effects in the 1H NMR signal of pyridinium H2; therefore, there are not expected strong differences between the behavior of this IL and that used as reference [1B3MPy][DCA]. Chart 1 depicts the ion-paired structures of studied ILs, Chart 1. Optimized Structures of the Different Ion Pairs at the B88-P86/TZVP Computational Level and σ-Surfaces Obtained by COSMO-RS

Figure 1. (A) σ-Profiles and (B) σ-potentials of different cations and the DCA anion obtained by COSMO-RS using the independent ions molecular model.

(above the cut off σH‑Bond > 0.0085 e/Å2); therefore, [DCA]− can be considered as a hydrogen bond acceptor (the σ-surface of Figure 1A reveals that the basic groups − red colored − are the central nitrogen and cyano nitrogen atoms). The pyridinium cation presents distribution of charge densities around zero (−0.0085 e/Å2 < σ < 0.0085 e/Å2), corresponding to the nonpolar alkyl and aromatic groups. In addition, all pyridinium cations show unresolved peaks of low intensity at values below the cut off σH‑Bond < −0.0085 e/Å2. As it can be seen, the polarized charge of these acidic groups is displaced toward a more negative polar region in the order −NH2 > −CN > −Cl > −CH3. These signals are mainly associated with the C2/H-2 hydrogen bond donor group of the pyridine (intensively blue colored in the σ-surface of the cation in Figure 1A) and, for the [1B3NH2Py]+ case, also with the hydrogens of the −NH2 group. The COSMO-RS method also provided the σ-potential of the ions (Figure 1B), which describes their interaction energy with compounds with a charge density [pX(σ)] with polarity σ. Attending COSMO-RS theory, [DCA]− will present attractive interactions with acidic groups and repellent ones with basic groups. In contrast, the pyridinium cation will promote exothermic and endothermic

optimized at the B88-P86/TZVP quantum-chemical approach. The calculated cation−anion distance, illustrated by the C2− H2···N(CN)2 hydrogen bond length (Table 1), supports that the cation−anion interaction strength is stronger in [1B3CNPy][DCA] and [1B3ClPy][DCA] than in [1B3MPy][DCA] and [1B3M5M2PPy][DCA]. In the case of [1B3NH2Py][DCA], it should be noted that the [DCA] anion interacts more favorably with hydrogen atoms of the −NH2 group than with H-2 of the pyridinium headgroup, which confers an additional acidic character to the [1B3NH2Py] cation. COSMO-RS Analysis of IL Solvent Properties. The RS method is a valuable computational tool to analyze the thermodynamic behavior of fluid mixtures containing ILs and aromatic compounds.47−51 COSMO-RS provides the polarized charge distribution (σ, sigma) of the individual compounds, easily visualized by the σ-profile histogram, whose qualitative analysis allows anticipating its behavior as a component in a mixture. Figure 1A collects the σ-profile of the independent [DCA] anion and [1B3MPy], [1B3M5M2PPy], [1B3ClPy], [1B3NH2Py], and [1B3CNPy] cations. This anion presents a characteristic peak at 0.016 e/Å2, the high polarity region 2018

DOI: 10.1021/acssuschemeng.6b02922 ACS Sustainable Chem. Eng. 2017, 5, 2015−2025

Research Article

ACS Sustainable Chemistry & Engineering mixtures with compounds presenting, respectively, hydrogen bond acceptor and donor groups. The σ-potentials clearly indicated that the functionalization of pyridinium ring increases the acidic character of the cation in the order [1B3NH2Py] > [1B3CNPy] > [1B3ClPy] > [1B3MPy] ∼ [1B3M5M2PPy]. These results are consistent with the cation−anion hydrogen bond strength revealed above from spectroscopic and computational molecular data. Previous COSMO-RS studies found in bibliography revealed that a molecular model of ion-pair (CA) for IL (as those optimized structures in Chart 1) may provide more accurate predictions of solvent properties than the model of independent ions (C+A).51−54 The analysis of the σ-profiles, σ-potentials, and σ-surface of the ion-paired IL structures (Figure S17 in the Supporting Information) provides a similar description of the studied ILs: amphoteric solvents able to efficiently solvate polar acidic compounds (thanks to the hydrogen bond nature of [DCA] anion) and, in a minor extent, polar basic compounds (thanks to the increasing acidic character of C2/H-2 in the series [1B3CNPy] > [1B3ClPy] > [1B3MPy] ∼ [1B3M5M2PPy] and the hydrogens of −NH2 in the [1B3NH2Py] cation). The aim of this work was to evaluate the role of the functionalization of the pyridinium ring on the IL ability to separate N-aromatic compounds from fuel mixtures. For this purpose, COSMO-RS was applied to analyze the mixing properties of the pyridine-IL and heptane-IL binary systems. Figure 2 presents the σ-profile, σ-potential, and σ-surface of the pyridine and heptane molecules. Based on COSMO-RS information, heptane is a nonpolar compound (with its polarized charge located at −0.0085 e/Å2 < σ < 0.0085 e/Å2 region), presenting strong repulsive interactions with both positive and negative polar groups. Pyridine also presents strong nonpolar fragments associated with its aromatic ring; however, it also shows a polarized charge above the cut off σH‑Bond > 0.0085 e/Å2, corresponding to the nitrogen group of the heterocyclic, which presents a strong hydrogen bond acceptor character. As a result, pyridine will present attractive interactions with acidic species as functionalized pyridinium cations. A quantitative analysis of the behavior of pyridine-IL and heptane-IL mixtures can be achieved predicting the excess enthalpy (HE), entropy (SE), and free energy (GE) by COSMO-RS (Figures 3 and 4). In addition, COSMO-RS provides the contribution of the different intermolecular interactions (hydrogen bond, misfit-electrostatic, and van der Waals) to the excess enthalpy, contributing to understanding the thermodynamics of the system from a microscopic point of view.52,55,56 Clearly, heptane presents no favorable mixtures with the studied ILs, being a phenomena controlled by the endothermic mixing enthalpy (Figure 3), consistently with the high immiscibility of these mixtures. The repulsive heptane-IL interactions become stronger with the inclusion of polar functional groups in the pyridinium ring, increasing in the order [1B3CNPy]+ > [1B3ClPy]+ > [1B3MPy]+ > [1B3,5M2PPy]+ (Figure 3A). In contrast, the systems formed by pyridine and functionalized pyridinium-based ILs present a very different behavior. Exothermic mixtures are obtained when increasing the acidic character of the cation by functionalizing with adequate functional groups (Figure 4A). It should be remarked that, in general, the mixtures of aromatic compounds and ILs are endothermic, governed by repulsive electrostatic interactions.47,49 The favorable hydrogen bonding explains the high solubility of the N-aromatic compounds in pyridinium-based

Figure 2. (A) σ-Profiles and (B) σ-potentials of n-heptane and pyridine obtained by COSMO-RS.

ILs comparing to other aromatic compounds.33 Therefore, the increased acidity of functionalized ILs [1B3CNPy][DCA] 3d, [1B3NH2Py][DCA] 3c, and [1B3ClPy][DCA] 3b would anticipate improved solvent properties for extracting pyridine solute. However, unexpectedly, COSMO-RS calculations reveal that the mixtures of pyridine and each of these ILs are phenomena driven by the decreasing entropy (Figure 4A). This behavior may unfavorably affect the suitability of new synthesized ILs as extracting agents in pyridine separation from aliphatic compounds. In contrast, the tetraalkyl substituted pyridinium based IL [1 B3 M 5 M 2 PPy][DCA] presents negative values of excess free energy, mainly due to the entropy increase in the mixture with pyridine (Figure 4A). Liquid−Liquid Equilibrium of the Pyridine-Heptane-IL System: Extracting Ability Analysis. The ability of the synthesized ILs as solvents for the extraction of pyridine from heptane was determined by measuring and predicting with COSMO-RS, the tie-lines covering the relevant area of the immiscibility region of the ternary systems (IL + heptane + pyridine). The experimental and calculated LLE data for the ternary systems are given in Tables 2−6. The corresponding triangular diagrams are plotted in Figure S18 of the Supporting Information. The obtained LLE data shows a favorable 2019

DOI: 10.1021/acssuschemeng.6b02922 ACS Sustainable Chem. Eng. 2017, 5, 2015−2025

Research Article

ACS Sustainable Chemistry & Engineering

Figure 3. (A) Description of the cation effect on HE, SE, and GE values of heptane−IL equimolar computed by COSMO-RS at T = 298 K. (B) Description of the cation effect on the excess enthalpy of heptane−IL equimolar mixtures in terms of the intermolecular interaction contributions [HE (MF), HE (HB), and HE (vdW)] computed by COSMO-RS at T = 298.

thermodynamic efficiency parameters for liquid−liquid extraction separation. These parameters were calculated as

extractive capacity of the ILs (higher pyridine molar fraction in the extract phase than in raffinate), low solubility of the aliphatic compound in the IL-reach phase (i.e., high selectivity for pyridine solute), and nondetectable solubility of the IL solvent in the aliphatic-reach phase. As a consequence, the IL concentration in the aliphatic product is nearly negligible, as it was recently demonstrated in a similar aromatic−aliphatic ILbased separation by process simulation studies.24,25 Figure 5 shows that COSMO-RS provides reasonable predictions of LLE data for both phases in the whole range of compositions, with mean error deviation (MEA) lower that 4%. Partition coefficients (β, eq 1) of solute between IL-rich and aliphaticrich phases and selectivity (S, eq 2) are commonly analyzed as

β=

S=

x 2II x 2I

(1)

x 2IIx1I x 2Ix1II

xI1

(2)

xI2

where and are the mole fractions of heptane and pyridine, respectively, in the alkane-rich phase (upper phase); and xII1 and xII2 are the mole fractions of heptane and pyridine, respectively, in the IL-rich phase (lower phase). Tables 2−6 collect the estimated β and S values from experimental and calculated tie-line compositions. For compar2020

DOI: 10.1021/acssuschemeng.6b02922 ACS Sustainable Chem. Eng. 2017, 5, 2015−2025

Research Article

ACS Sustainable Chemistry & Engineering

Figure 4. (A) Description of the cation effect on HE, SE, and GE values of pyridine−IL equimolar computed by COSMO-RS at T = 298 K. (B) Description of the cation effect on the excess enthalpy of pyridine−IL equimolar mixtures in terms of the intermolecular interaction contributions [HE (MF), HE (HB), and HE (vdW)] computed by COSMO-RS at T = 298 K.

Table 2. Experimental Liquid−Liquid Equilibrium Data in Mole Fraction and Experimental and Calculated Values of the Solute Distribution Ratio (β) and Selectivity (S) for the Ternary Systems (Heptane (1) + Pyridine (2) + [1B3MPy][DCA] (3)) at T = 298.15 K and Atmospheric Pressure

Table 3. Experimental Liquid−Liquid Equilibrium Data in Mole Fraction and Experimental and Calculated Values of the Solute Distribution Ratio (β) and Selectivity (S) for the Ternary Systems (Heptane (1) + Pyridine (2) + [1B3ClPy][DCA] (3)) at T = 298.15 K and Atmospheric Pressure

alkane rich phase (I)

alkane rich phase (I)

ionic liquid rich phase (II)

exp

calc

ionic liquid rich phase (II)

exp

calc

x1

x2

x1

x2

β

S

β

S

x1

x2

x1

x2

β

S

β

S

0.986 0.967 0.955 0.947 0.923

0.014 0.033 0.045 0.053 0.077

0.016 0.014 0.014 0.014 0.013

0.112 0.208 0.281 0.317 0.380

8.00 6.30 6.24 5.98 4.94

493.00 435.36 425.96 404.58 350.39

5.58 5.34 5.15 5.05 4.82

260.62 233.60 212.08 197.55 177.78

0.987 0.973 0.961 0.934 0.890 0.839

0.013 0.027 0.039 0.066 0.110 0.161

0.013 0.012 0.011 0.011 0.012 0.011

0.102 0.179 0.242 0.346 0.464 0.564

7.85 6.63 6.21 5.24 4.22 3.50

595.70 537.55 542.10 445.13 312.85 267.19

4.85 4.78 4.55 4.25 3.96 3.69

528.11 460.58 392.46 326.43 233.19 173.46

2021

DOI: 10.1021/acssuschemeng.6b02922 ACS Sustainable Chem. Eng. 2017, 5, 2015−2025

Research Article

ACS Sustainable Chemistry & Engineering Table 4. Experimental Liquid−Liquid Equilibrium Data in Mole Fraction and Experimental and Calculated Values of the Solute Distribution Ratio (β) and Selectivity (S) for the Ternary Systems (Heptane (1) + Pyridine (2) + [1B3NH2Py][DCA] (3)) at T = 298.15 K and Atmospheric Pressure alkane rich phase (I) x1

x2

x1

x2

β

S

β

S

0.984 0.974 0.968 0.943 0.904 0.858

0.016 0.026 0.032 0.057 0.096 0.142

0.016 0.016 0.015 0.014 0.015 0.016

0.120 0.204 0.218 0.371 0.483 0.581

7.50 7.85 6.81 6.51 5.03 4.09

461.25 477.63 439.63 438.41 303.22 219.41

5.52 5.30 5.17 4.79 4.45 4.14

491.6 463.8 413.2 316.3 233.7 169.7

Table 5. Experimental Liquid−Liquid Equilibrium Data in Mole Fraction and Experimental and Calculated Values of the Solute Distribution Ratio (β) and Selectivity (S) for the Ternary Systems (Heptane (1) + Pyridine (2) + [1B3CNPy][DCA] (3)) at T = 298.15 K and Atmospheric Pressure alkane rich phase (I)

ionic liquid rich phase (II)

exp

Figure 5. Experimental vs COSMO-RS calculated LLE compositions of heptane-pyridine-IL systems.

mixtures (Figure 4B). At more concentrated pyridine mixtures, the partition coefficient value decreases; obtaining closer results when using the different functionalized pyridinium-based ILs. On the other hand, all studied ILs present remarkably higher selectivity for pyridine with respect to the heptane compound. In this case, the increasing acidity of the pyridinium cation by incorporation of polar functional groups generally improves the extracting properties of ILs (S increasing in the order [1B3M5M2PPy][DCA] < [1B3MPy][DCA] 3a < [1B3NH2Py][DCA] 3c < [1B3ClPy][DCA] 3b), since it promotes lower solubility of heptane in ILs. As expected, the selectivity of ILs decreases with the content of pyridine in the mixture, and the solvent behavior differences between the studied ILs are minor. Experimental S values for [1B3CNPy][DCA] 3d solvent are deviated from the general trend and from COSMO-RS calculations, probably due to the difficulties found in LLE measurements. For this kind of systems, the compositions of heptane in the IL-rich are very low (see Tables 2−6), and a small change in these values strongly affects the selectivity values. For that, the experimental S values reported in this work should be interpreted as a range, especially in those systems with the lowest values.

calc

x1

x2

x1

x2

β

S

β

S

0.985 0.970 0.941 0.934 0.849

0.015 0.030 0.059 0.066 0.151

0.018 0.016 0.014 0.012 0.017

0.098 0.173 0.251 0.267 0.423

6.53 5.77 4.25 4.05 2.80

357.52 349.60 285.95 314.87 139.90

3.42 3.31 3.16 3.11 2.85

1667.04 1576.75 1458.95 1427.38 808.44

Table 6. Experimental Liquid−Liquid Equilibrium Data in Mole Fraction and Experimental from Ref 11 and Calculated Values of the Solute Distribution Ratio (β) and Selectivity (S) for the Ternary Systems (Heptane (1) + Pyridine (2) + [1B3M5M2PPy][DCA] (3)) at T = 298.15 K and Atmospheric Pressure alkane rich phase (I)

ionic liquid rich phase (II)

exp

calc

x1

x2

x1

x2

β

S

β

S

0.992 0.982 0.973 0.962

0.008 0.018 0.027 0.038

0.082 0.059 0.060 0.057

0.112 0.208 0.263 0.319

14.00 11.56 9.74 8.39

169.37 192.33 157.96 141.68

7.92 7.29 7.23 6.85

37.03 34.51 34.59 32.98



CONCLUSIONS Different functional groups (X = cyano, amino, choro, alkyl) were added to the 1-butyl-3-Xpyridinium cation, with the aim of synthesizing new dicyanamide-based ILs with modified solvent properties. Polarity and hydrogen bond donor character of these ILs were found shifted by the functional group nature, increasing in the order [1B3CNPy][DCA] > [1B3ClPy][DCA] > [1B3NH2Py][DCA] > [1B3MPy][DCA] ∼ [1B3,5 M2PPy][DCA], on the basis of experimental and computational analysis. COSMO-RS methodology was applied to evaluate the excess properties of heptane-IL and pyridine-IL binary mixtures. It was obtained that the mixing phenomena were enthalpically and entropically unfavorable for heptane-IL mixtures. In contrast, attractive hydrogen bond interactions occur in pyridine-IL mixtures. It is also possible to favor the thermodynamics of the mixing phenomena by using polyalkylsubstituted pyridinium cations, increasing the entropy of the system. Liquid−liquid equilibrium measurements and COSMO-RS calculations demonstrated that the new function-

ison purposes, β and S values for [1B3MPy][DCA] 3a, [1B3ClPy][DCA] 3b, [1B3NH2Py][DCA] 3c, [1B3CNPy][DCA] 3d, and [1B3,5M2PPy][DCA] were estimated at fixed pyridine composition in the heptane-rich phase of XI2 = 0.02 and 0.10 (Figures 6 and 7), by interpolation (see Table S2 in the Supporting Information) from tie-lines LLE data in Tables 2−6. As can be seen, all pyridinium-based [DCA] ILs present high partition coefficients (β ≫1), confirming their adequate solvation properties for pyridine extraction. The separation capacity is clearly higher for the tetraalkyl substituted pyridinium based IL [1B3M5M2PPy][DCA], indicating that liquid−liquid equilibrium is controlled by entropic effects. Consistently, the β value slightly decreases in the order [1B3M5M2PPy][DCA] > [1B3MPy][DCA] 3a > [1B3NH2Py][DCA] 3c > [1B3ClPy][DCA] 3b > [1B3CNPy][DCA] 3d, in agreement with the decreasing entropy of pyridine-IL binary 2022

DOI: 10.1021/acssuschemeng.6b02922 ACS Sustainable Chem. Eng. 2017, 5, 2015−2025

Research Article

ACS Sustainable Chemistry & Engineering

Figure 6. Partition coefficient (β) for different pyridine compositions in the heptane-rich phase (XI2). Experimental and COSMO-RS results.

Figure 7. Selectivity (S) for different pyridine compositions in the heptane-rich phase (XI2). Experimental and COSMO-RS results.



alized 1-butyl-3-Xpyridinium dicyanamide (X = cyano, amino, choro, alkyl) ILs present good performance as extracting agents of N-compounds from fuel. High partition coefficients and selectivity for pyridine extraction from heptane were achieved, being the separation improved when using tetraalkyl substituted pyridinium ILs, due to favorable entropic effects.



AUTHOR INFORMATION

Corresponding Author

*Phone: +34 986812290. E-mail: [email protected]. ORCID

Emilia Tojo: 0000-0002-7099-3437 Author Contributions

ASSOCIATED CONTENT

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acssuschemeng.6b02922. Detailed experimental synthesis and characterization data of the synthesized ILs; figures with the σ-profiles and σpotentials of cations and anions obtained by COSMORS; triangular diagrams with the experimental tie-lines for the ternary mixtures {heptane (1) + pyridine (2) + IL (3)}; table with the calculated parameters resonance effect (σR+), inductive field effect (σF)e, and polarizability effect (σα) for the studied cations; and table with the interpolation of β (solute distribution ratio) and S (selectivity) values for the studied ILs at fixed pyridine composition (PDF)

Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS

We thank the Xunta de Galicia (REGALIs Network R2014/ 015) and Comunidad de Madrid (project S2013-MAE-2800) for their financial support. We are also grateful to the “Centro ́ de Computación Cientifica de la Universidad Autónoma de Madrid” for computational facilities as well as the research support services of the University of Vigo (CACTI) for the NMR and MS divisions work. 2023

DOI: 10.1021/acssuschemeng.6b02922 ACS Sustainable Chem. Eng. 2017, 5, 2015−2025

Research Article

ACS Sustainable Chemistry & Engineering



K: Experiments and correlations. Fluid Phase Equilib. 2012, 324, 17− 27. (20) Shah, M. R.; Anantharaj, R.; Banerjee, T.; Yadav, G. D. Quaternary (liquid + liquid) equilibria for systems of imidazolium based ionic liquid + thiophene + pyridine + cyclohexane at 298.15 K: Experiments and quantum chemical predictions. J. Chem. Thermodyn. 2013, 62, 142−150. (21) Abro, R.; Abro, M.; Gao, S.; Bhutto, A. W.; Ali, Z. M.; Shah, A.; Chen, X.; Yu, G. Extractive denitrogenation of fuel oils using ionic liquids: a review. RSC Adv. 2016, 6, 93932−93946. (22) Chen, X.; Yuan, S.; Abdeltawab, A. A.; Al-Deyab, S. S.; Zhang, J.; Yu, L.; Yu, G. Extractive desulfurization and denitrogenation of fuels using functional acidic ionic liquids. Sep. Purif. Technol. 2014, 133, 187−193. (23) Jie, L.; Bo, M. Removal of Nitrogen Compounds from Shale Diesel Fraction Using Ionic Liquid [C4mim]HSO4. China Pet. Process. Petrochem. Technol. 2016, 18 (3), 15−21. (24) De Riva, J.; Ferro, V. R.; Moreno, D.; Diaz, I.; Palomar, J. Aspen Plus supported conceptual design of the aromatic−aliphatic separation from low aromatic content naphtha using 4-methyl-N-butylpyridinium tetrafluoroborate ionic liquid. Fuel Process. Technol. 2016, 146, 29−38. (25) Ferro, V. R.; De Riva, J.; Sanchez, D.; Ruiz, E.; Palomar, J. Conceptual design of unit operations to separate aromatic hydrocarbons from naphtha using ionic liquids. COSMO-based process simulations with multi-component “real” mixture feed. Chem. Eng. Res. Des. 2015, 94, 632−647. (26) Chen, S.; Vijayaraghavan, R.; MacFarlane, D. R.; Izgorodina, E. I. Ab Initio Prediction of Proton NMR Chemical Shifts in Imidazolium Ionic Liquids. J. Phys. Chem. B 2013, 117 (11), 3186−3197. (27) Hizaddin, H. F.; Hadj-Kali, M. K.; Ramalingam, A.; Hashim, M. A. Extraction of nitrogen compounds from diesel fuel using imidazolium- and pyridinium-based ionic liquids: Experiments, COSMO-RS prediction and NRTL correlation. Fluid Phase Equilib. 2015, 405, 55−67. (28) Królikowska, M.; Karpińska, M. Extraction of aromatic nitrogen compounds from heptane using quinolinium and isoquinolinium based ionic liquids. Fluid Phase Equilib. 2015, 400, 1−7. (29) Domanska, U.; Lukoshko, E. V. Separation of pyridine from heptane with tricyanomethanide-based ionic liquids. Fluid Phase Equilib. 2015, 395, 9−14. (30) Lukoshko, E.; Mutelet, F.; Paduszyński, K.; Domańska, U. Phase diagrams of binary systems containing tricyanomethanide-based ionic liquids and thiophene or pyridineNew experimental data and PCSAFT modelling. Fluid Phase Equilib. 2015, 399, 105−114. (31) Vilas, M.; González, E. J.; Tojo, E. Extractive denitrogenation of model oils with tetraalkyl substituted pyridinium based ionic liquids. Fluid Phase Equilib. 2015, 396, 66−73. (32) Rogošić, M.; Sander, A.; Kojić, V.; Vuković, J. P. Liquid−liquid equilibria in the ternary and multicomponent systems involving hydrocarbons, thiophene or pyridine and ionic liquid (1-benzyl-3metylimidazolium bis(trifluorometylsulfonyl)imide. Fluid Phase Equilib. 2016, 412, 39−50. (33) Wlazlo, M.; Karpinska, M.; Domanska, U. A 1-alkylcyanopyridinium-based ionic liquid in the separation processes. J. Chem. Thermodyn. 2016, 97, 253−260. (34) Lü, R.; Qu, Z.; Yu, H.; Wang, F.; Wang, S. Comparative study on interactions between ionic liquids and pyridine/hexane. Chem. Phys. Lett. 2012, 532, 13−18. (35) Lü, R.; Qu, Z.; Lin, J. Comparative study of interactions between thiophene\pyridine\benzene\heptane and 1-butyl-3-methylimidazolium trifluoromethanesulfonate by density functional theory. J. Mol. Liq. 2013, 180, 207−214. (36) Hizaddin, H. F.; Hashim, M. A.; Anantharaj, R. Evaluation of Molecular Interaction in Binary Mixture of Ionic Liquids + Heterocyclic Nitrogen Compounds: Ab Initio Method and COSMO-RS Model. Ind. Eng. Chem. Res. 2013, 52 (50), 18043− 18058.

ABBREVIATIONS NMR, nuclear magnetic resonance; HB, hydrogen bond; COSMO-RS, conductor-like screening model for real solvents; IL, ionic liquid; LLE, liquid−liquid equilibrium; HDN, hydrodenitrogenation; ICP-MS, inductively coupled plasma mass spectrometry; HRMS, high resolution mass spectrometry; MEA, mean error deviation



REFERENCES

(1) United States Environment Protection Agency (EPA). https:// www.epa.gov/emission-standards-reference-guide (accessed Dec 22, 2016). (2) DieselNet. https://www.dieselnet.com/standards/eu/ld.php/ (accessed Dec 22, 2016). (3) Shin, S.; Sakanishi, K.; Mochida, I.; Grudoski, D. A.; Shinn, J. H. Identification and Reactivity of Nitrogen Molecular Species in Gas Oils. Energy Fuels 2000, 14 (3), 539−544. (4) Wiwel, P.; Knudsen, K.; Zeuthen, P.; Whitehurst, D. Assessing Compositional Changes of Nitrogen Compounds during Hydrotreating of Typical Diesel Range Gas Oils Using a Novel Preconcentration Technique Coupled with Gas Chromatography and Atomic Emission Detection. Ind. Eng. Chem. Res. 2000, 39 (2), 533−540. (5) Babich, I. V.; Moulijn, J. A. Science and technology of novel processes for deep desulfurization of oil refinery streams: a review. Fuel 2003, 82 (6), 607−631. (6) Caeiro, G.; Costa, A. F.; Cerqueira, H. S.; Magnoux, P.; Lopes, J. M.; Matias, P.; Ribeiro, F. R. Nitrogen poisoning effect on the catalytic cracking of gasoil. Appl. Catal., A 2007, 320, 8−15. (7) Eßer, J.; Wasserscheid, P.; Jess, A. Deep desulfurization of oil refinery streams by extraction with ionic liquids. Green Chem. 2004, 6, 316−322. (8) Gabric, B.; Sander, A.; Cvjetko Bubalo, M.; Macut, D. Extraction of S- and N-Compounds from the Mixture of Hydrocarbons by Ionic Liquids as Selective Solvents. Sci. World J. 2013, 2013, 1−11. (9) Anantharaj, R.; Banerjee, T. COSMO-RS-Based Screening of Ionic Liquids as Green Solvents in Denitrification Studies. Ind. Eng. Chem. Res. 2010, 49 (18), 8705−8725. (10) Huddleston, J. G.; Willauer, H. D.; Swatloski, R. P.; Visser, A. E.; Rogers, R. D. Room temperature ionic liquids as novel media for ‘clean’ liquid−liquid extraction. Chem. Commun. 1998, 1765−1766. (11) Perez De Los Rios, A.; Hernandez Fernandez, F. J. Ionic Liquids in Separation Technology; Amsterdam, The Netherlands, 2014. (12) Rodríguez, H. Ionic Liquids for Better Separation Processes; Springer-Verlag: Berlin, Heidelberg, 2016. (13) Anantharaj, R.; Banerjee, T. Evaluation and comparison of global scalar properties for the simultaneous interaction of ionic liquids with thiophene and pyridine. Fluid Phase Equilib. 2010, 293 (1), 22− 31. (14) Anantharaj, R.; Banerjee, T. Quantum chemical studies on the simultaneous interaction of thiophene and pyridine with ionic liquid. AIChE J. 2011, 57 (3), 749−764. (15) Anantharaj, R.; Banerjee, T. Fast Solvent Screening for the Simultaneous Hydrodesulfurization and Hydrodenitrification of Diesel Oil Using Ionic Liquids. J. Chem. Eng. Data 2011, 56 (6), 2770−2785. (16) Anantharaj, R.; Banerjee, T. Liquid−liquid equilibria for quaternary systems of imidazolium based ionic liquid + thiophene + pyridine + iso-octane at 298.15 K: Experiments and quantum chemical predictions. Fluid Phase Equilib. 2011, 312, 20−30. (17) Anantharaj, R.; Banerjee, T. Liquid−Liquid Equilibrium Studies on the Removal of Thiophene and Pyridine from Pentane Using Imidazolium-Based Ionic Liquids. J. Chem. Eng. Data 2013, 58 (4), 829−837. (18) Hansmeier, A. R.; Meindersma, G. W.; de Haan, A. B. Desulfurization and denitrogenation of gasoline and diesel fuels by means of ionic liquids. Green Chem. 2011, 13, 1907−1913. (19) Ravilla, U. K.; Banerjee, T. Liquid liquid equilibria of imidazolium based ionic liquid + pyridine + hydrocarbon at 298.15 2024

DOI: 10.1021/acssuschemeng.6b02922 ACS Sustainable Chem. Eng. 2017, 5, 2015−2025

Research Article

ACS Sustainable Chemistry & Engineering (37) Hizaddin, H. F.; Anantharaj, R.; Hashim, M. A. A quantum chemical study on the molecular interaction between pyrrole and ionic liquids. J. Mol. Liq. 2014, 194, 20−29. (38) Diedenhofen, M.; Klamt, A. COSMO-RS as a tool for property prediction of IL mixturesA review. Fluid Phase Equilib. 2010, 294, 31−38. (39) Anantharaj, R.; Banerjee, T. Aromatic sulfur-nitrogen extraction using ionic liquids: Experiments and predictions using an a priori model. AIChE J. 2013, 59 (12), 4806−4815. (40) TURBOMOLE V7.0; TURBOMOLE GmbH: 2015. (41) Frisch, M. J.; Schlegel, H. B.; Scuseria, G. E. Gaussian 09, Revision E.01; Gaussian, Inc: Wallingford, CT, 2013. (42) Eckert, F.; Klamt, A. Fast solvent screening via quantum chemistry: COSMO-RS approach. AIChE J. 2002, 48 (2), 369−385. (43) Eckert, F.; Klamt, A. COSMOtherm Version C3.0, Release 16.01; COSMOlogic GmbH & Co. KG: Leverkusen, Germany, 2015. (44) Verdía, P.; González, E. J.; Rodríguez-Cabo, B.; Tojo, E. Synthesis and characterization of new polysubstituted pyridiniumbased ionic liquids: application as solvents on desulfurization of fuel oils. Green Chem. 2011, 13 (10), 2768−2776. (45) Bankmann, D.; Giernoth, R. Magnetic resonance spectroscopy in ionic liquids. Prog. Nucl. Magn. Reson. Spectrosc. 2007, 51 (1), 63− 90. (46) Palomar, J.; Ferro, V. R.; Gilarranz, M. A.; Rodríguez, J. J. Computational Approach to Nuclear Magnetic Resonance in 1-Alkyl3-methylimidazolium Ionic Liquids. J. Phys. Chem. B 2007, 111 (1), 168−180. (47) Calvar, N.; Domínguez, I.; Gómez, E.; Palomar, J.; Domínguez, A. Evaluation of ionic liquids as solvent for aromatic extraction: Experimental, correlation and COSMO-RS predictions. J. Chem. Thermodyn. 2013, 67, 5−12. (48) Bedia, J.; Ruiz, E.; de Riva, J.; Ferro, V. R.; Palomar, J.; Rodríguez, J. J. Optimized ionic liquids for toluene absorption. AIChE J. 2013, 59 (5), 1648−1656. (49) Domínguez, I.; González, E. J.; Palomar, J.; Domínguez, A. Phase behavior of ternary mixtures {aliphatic hydrocarbon + aromatic hydrocarbon + ionic liquid}: Experimental LLE data and their modeling by COSMO-RS. J. Chem. Thermodyn. 2014, 77, 222−229. (50) Gómez, E. L.; Domínguez, I.; Calvar, N.; Palomar, J.; Domínguez, A. Experimental data, correlation and prediction of the extraction of benzene from cyclic hydrocarbons using [Epy][ESO4] ionic liquid. Fluid Phase Equilib. 2014, 361, 83−92. (51) Ferro, V. R.; de Riva, J.; Sanchez, D.; Ruiz, E.; Palomar, J. Conceptual design of unit operations to separate aromatic hydrocarbons from naphtha using ionic liquids. COSMO-based process simulations with multi-component “real” mixture feed. Chemical Engineering Research & Design 2015, 94, 632−647. (52) Ruiz, E.; Ferro, V. R.; Palomar, J.; Ortega, J.; Rodríguez, J. J. Interactions of Ionic Liquids and Acetone: Thermodynamic Properties, Quantum-Chemical Calculations, and NMR Analysis. J. Phys. Chem. B 2013, 117 (24), 7388−7398. (53) González-Miquel, M.; Palomar, J.; Rodríguez, F. Selection of Ionic Liquids for Enhancing the Gas Solubility of Volatile Organic Compounds. J. Phys. Chem. B 2013, 117 (1), 296−306. (54) Omar, S.; Lemus, J.; Ruiz, E.; Ferro, V. R.; Ortega, J.; Palomar, J. Ionic Liquid MixturesAn Analysis of Their Mutual Miscibility. J. Phys. Chem. B 2014, 118 (9), 2442−2450. (55) Navas, A.; Ortega, J.; Vreekamp, R.; Marrero, E.; Palomar, J. Experimental Thermodynamic Properties of 1-Butyl-2-methylpyridinium Tetrafluoroborate [b2mpy][BF4] with Water and with Alkan-1-ol and Their Interpretation with the COSMO-RS Methodology. Ind. Eng. Chem. Res. 2009, 48 (5), 2678−2690. (56) Vreekamp, R.; Castellano, D.; Palomar, J.; Ortega, J.; Espiau, F.; Fernández, L.; Penco, E. Thermodynamic Behavior of the Binaries 1Butylpyridinium Tetrafluoroborate with Water and Alkanols: Their Interpretation Using 1H NMR Spectroscopy and Quantum-Chemistry Calculations. J. Phys. Chem. B 2011, 115 (27), 8763−8774.

2025

DOI: 10.1021/acssuschemeng.6b02922 ACS Sustainable Chem. Eng. 2017, 5, 2015−2025