J. Chem.

Thermodynamics 46 (2012) 2–28

Contents lists available at ScienceDirect

J. Chem. Thermodynamics
journal homepage: www.elsevier.com/locate/jct

Review

Ionic liquids in separations of azeotropic systems – A review

A.B. Pereiro ⇑, J.M.M. Araújo, J.M.S.S. Esperança, I.M. Marrucho, L.P.N. Rebelo ⇑
Instituto de Tecnologia Química e Biológica, Universidade Nova de Lisboa, Apartado 127, 2780-901 Oeiras, Portugal

a r t i c l e i n f o a b s t r a c t

Article history: Efforts to make existing separation methods more efficient and eco-friendly may get a boost from the use
Available online 30 May 2011 of a relatively new class of compounds known as ionic liquids (ILs). The separation of azeotropic mixtures
has conventionally been one of the most challenging tasks in industrial processes due to the fact that
Keywords: their separation by simple distillation is basically impossible.
Ionic liquids This paper provides a critical review of methods using ILs as azeotrope breakers. Three separation pro-
Azeotropes cesses were addressed: liquid–liquid extraction, extractive distillation, and supported liquid membranes.
Separation
We examine the azeotrope breaking potential of ILs and compare their performance to that of conven-
Extraction
Distillation
tional solvents. A systematic analysis of the influence of the structure of ILs on their azeotrope breaking
Membranes capacity contributes to the establishment of guidelines for selecting the most suitable ILs for the separa-
tion of specific azeotropic mixtures.
Ó 2011 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 03
1.1. Azeotropes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 03
1.2. Ionic liquids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 04
1.3. Energy and economic costs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 05
1.4. Separation process. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 06
1.4.1. Extractive or azeotropic distillation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 06
1.4.2. Liquid–Liquid extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 07
1.4.3. Supported liquid membranes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 07
2. Azeotrope separation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 07
2.1. Aqueous systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 07
2.1.1. (Water + ethanol) system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 07
2.1.1.1. Separation by extractive distillation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 07
2.1.1.2. Separation by other methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.1.2. (Water + other alcohols) systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.1.2.1. Separation by extractive distillation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.1.2.2. Separation by other methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.1.3. (Water + tetrahydrofuran) system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.1.3.1. Separation by extractive distillation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.1.3.2. Separation by liquid–liquid extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2. Alcoholic systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.1. (Alcohols + ester) systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.2. (Alcohols + ketones) systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.2.3. (Alcohols + aliphatic hydrocarbons) systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.2.4. (Alcohols + halogenated) systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.3. (Aromatic + aliphatic hydrocarbons) systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.3.1. (Benzene + hexane) system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

⇑ Corresponding authors. Tel.: +351 214469414; fax: +351 214411277.

E-mail addresses: anab@itqb.unl.pt (A.B. Pereiro), luis.rebelo@itqb.unl.pt (L.P.N. Rebelo).
URL: www.itqb.unl.pt (A.B. Pereiro)

0021-9614/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jct.2011.05.026
 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28 3

2.3.1.1. Effect of the alkyl chain length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

2.3.1.2. Effect of the nature of the anion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.3.1.3. Effect of the nature of the cation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.3.2. (Benzene + heptane) system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.3.3. Other (aromatic + aliphatic hydrocarbons) systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.4. Other azeotropes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.4.1. (Methanol + dimethyl carbonate) system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.4.2. (Ethanol + ethyl tert-butyl ether) system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.4.3. (Hexane + ethyl acetate) and (cyclohexane + 2-butanone) systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.5. Final remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3. Conclusions and future work. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Appendix A. Supplementary data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

List of symbols

T temperature V Mix
Solute volume of solute in the initial azeotropic mixture (LLE
P pressure data)
az azeotrope EEP extraction efficiency from extraction process data
IL, ILs ionic liquid, ionic liquids wFeed
Inert mass fraction of the inert in feed stream (extraction pro-
LLE (liquid + liquid) equilibria cess data)
VLE (vapor + liquid) equilibria wRaffinate
Inert mass fraction of the inert in raffinate stream (extraction
aVLE
ij relative volatility (component i/component j) from (va- process data)
por + liquid) equilibria data wEquilibrium
Inert mass fraction of the inert in one equilibrium stage
yi, yj mole fraction of the component i or j in the vapor phase (theoretical data)
xi, xj mole fraction of the component i or j in the liquid phase SLMs, SILMs supported liquid membranes, supported ionic liquid
(for VLE data the mole fraction is ionic liquid free basis) membranes
SVLE
IL selectivity from (vapor + liquid) equilibria data SMEM selectivity from supported liquid membranes data
awith
ij
IL
relative volatility from VLE data using as entrainer ionic PermSolute permeability of the solute in supported liquid mem-
liquid branes
awithout
ij
IL
relative volatility from VLE data without ionic liquid PermInert permeability of the inert in supported liquid mem-
(azeotrope) branes
bLLE distribution coefficient from (liquid + liquid) equilibria aMEM separation factor from supported liquid membranes
data data
xIL Phase
Solute mole fraction of the solute in ionic liquid phase (LLE pPermeate
Solute partial pressure of the solute in the permeate
data) pPermeate partial pressure of the inert in the permeate
Inert
xOrganic Phase
mole fraction of the solute in organic phase (LLE
Solute pFeed partial pressure of the solute in the feed
data) Solute

SLLE selectivity from (liquid + liquid) equilibria data pFeed

Inert partial pressure of the inert in the feed
xIL Phase
Inert mole fraction of the inert in ionic liquid phase (LLE data) w, wt mass fraction
THF tetrahydrofuran
xOrganic
Inert
Phase
mole fraction of the inert in organic phase (LLE data)
t ton
ELLE extraction efficiency from (liquid + liquid) equilibria € euro
data h hour
C IL Phase
Solute concentration of solute in ionic liquid phase (LLE data) mol mole
V IL Phase
Solute volume of solute in ionic liquid phase (LLE data) n alkyl chain length
C Mix concentration of solute in the initial azeotropic mixture ETBE ethyl tert-butyl ether
Solute
(LLE data) C6, C7 hexane, heptane

1. Introduction ran, ter-methyl amyl ether, isopropanol, and vinyl acetate. Addi-
tionally, due to their growing popularity, biofuel processes
1.1. Azeotropes typically produce fermented products that form azeotropes with
water which is abundant in the fermentator. Among those prod-
Increasing concern about environmental issues, as well as the ucts, ethanol is the most important example, due to its excellent
establishment of new regulations, has recently directed the atten- properties as alternative fuel. However, compounds such as buta-
tion of the scientific community to novel processes based on green- nol also form aqueous azeotropic mixtures.
er technologies. In many areas of industry, solvent mixtures Several potential processes for separating azeotropes have
accumulate due to recycling difficulties. The separation of these come to light such as azeotropic distillation, extractive distillation,
mixtures into their pure components is necessary so that they pressure swing distillation, liquid–liquid extraction, adsorption,
can be reused. However, most of these solvent mixtures contain membranes. Extractive distillation is the most common process
azeotropes and, thus, their separation by simple distillation be- for removing one of the components in the azeotropic system. Its
comes impossible. Azeotropes are involved in common chemical greatest constraint is the tremendous energy costs required to
processes including the production of methyl acetate, terahydrofu- achieve a fluid phase system. Therefore, liquid–liquid separation,
4 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28

TABLE 1
Azeotropes studied using ionic liquids as separating agent (azeotrope data from reference [1]).

Component 1 Component 2 Component 3 Taz/K y1,az y2,az Paz/kPa Type

Aqueous systems
(Water + alcohols)
Water Ethanol 351.3 0.103 101.33 OX
Water Ethanol Ethyl acetate 343.4 0.09 0.084 101.33 OX
Water 2-Propanol 353.7 0.326 101.33 OX
Water 1-Propanol 360.8 0.568 101.33 OX
Water 2-Methyl-2-propanol 353.0 0.401 101.33 OX
Water 1-Butanol 365.5 0.754 101.33 EX
(Water + tetrahydrofuran)
Water Tetrahydrofuran 336.7 0.183 101.33 OX

Alcoholic systems
(Alcohols + esters)
Methanol Methyl acetate 328.2 0.348 107.19 OX
Ethanol Ethyl acetate 344.9 0.459 101.33 OX
1-Propanol 1-Propyl acetate 367.9 0.619 101.33 OX
2-Propanol Ethyl acetate 348.0 0.305 101.33 OX
1-Butanol 1-Butyl acetate 390.0 0.770 101.33 OX
(Alcohols + ketones)
Methanol Acetone 328.3 0.240 101.33 OX
Ethanol 2-Butanone 347.2 0.508 101.33 OX
2-Propanol 2-Butanone 350.6 0.383 101.33 OX
(Alcohols + aliphatics hydrocarbons)
Methanol Heptane 332.0 0.728 101.33 OX
Ethanol Hexane 331.7 0.341 101.33 OX
Ethanol Heptane 345.2 0.668 101.33 EX
(Alcohols + halogenated)
Methanol Chloroform 328.2 0.352 107.99 OX
Ethanol Chloroform 332.5 0.159 101.33 OX
(Aromatic + aliphatic hydrocarbons) systems
Benzene Hexane 341.5 0.050 101.33 OX
Benzene Cyclohexane 353.2 0.546 109.18 OX
Benzene Heptane 353.3 0.992 101.32 OX
Pyridine Heptane 368.6 0.300 101.33 OX
Other systems
Methanol Dimethyl carbonate 337.3 0.850 102.52 OX
Ethanol Ethyl tert-butyl ether 340.0 0.373 101.72 OX
Hexane Ethyl acetate 338.0 0.657 101.32 OX
Cyclohexane 2-Butanone 345.0 0.562 101.33 OX

O: homogeneous azeotrope in a completely miscible system.

E: heterogeneous azeotrope.
X: pressure maximum.

based on the immiscibility between two liquid phases at room media for engineers [2,3], not only due to their remarkable physi-
temperature, has emerged as a more economical and environmen- cochemical properties but also for their recyclability. Moreover, ILs
tally beneficial option. can be tailored for specific applications by accurately selecting the
The thermophysical properties of ionic liquids (ILs), as de- chemical nature of both the cation and the anion [4]. This fact,
scribed below, enable them to make important contributions to combined with the immense number of possible combinations of
azeotropic breaking. Table 1 presents a list of the systems in which cation and anion, have conferred upon them the designation of
ILs were used as separating agents together with the azeotropic ‘‘designer solvents’’. This feature is very attractive for chemical
point (T, P, y) coordinates [1]. engineers since the fine-tuning of solvent properties allows them
A recent upsurge in publications published until 2010 regarding to optimize process efficiency and cost.
azeotrope breaking using ILs has created the need for this review Ionic liquids were historically thought to be non-volatile, non-
article, which limits itself to considering direct methods. Despite flammable thermally, and chemically stable salts. However, based
their potential to provide valuable insights into the interactions on recent data, these assumptions have been progressively recon-
between the organic compounds and ionic liquids, important sidered [5,6]. The toxicity of ILs is also a hot issue that is driving the
sources of experimental data such as activity coefficients at infinite search for new environmentally friendly alternatives [7]. It is now
dilution and carbon dioxide solubility, as well as models (most commonly accepted that the label ‘‘green solvents’’ does not apply
commonly activity coefficients models, UNIQUAC, NRTL and COS- to all ILs in all situations. Nevertheless, a wide range of IL applica-
MO-RS) plus molecular simulation studies are beyond the scope tions regularly provide either better yields or lower toxicities than
of this work. conventional processes.
One of the relevant problems in the chemical industry is the
1.2. Ionic liquids separation of azeotropes. The remarkable azeotrope breaking
capacity of ILs, their almost null volatility at room temperature
Ionic liquids have been recognized for about a century, but have [5], and their recycling ease, have granted them recognition as pos-
only started receiving closer attention in the last two decades. They sible sustainable alternatives to conventional solvents. Moreover,
have recently become one of the increasingly popular ‘‘green’’ the physical and corrosion properties of ILs are generally more
 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28 5

FIGURE 1. Mapping of azeotropes in the matrix of ILs anions and cations used as separating agents. The colors refer to the number of different studied azeotropes from
literature per ion combination; the length of the bars gives the count per ion.

adequate for separation processes than those of inorganic molten and optimization of the distillation train have a critical impact on
salts, opening the door to their use in azeotropic separation the economics of the entire process [10].
processes. In extractive distillation, the entrainer does not need to be
Figure 1 maps the large number of publications already avail- vaporized and can be separated easily in a secondary distillation
able on this subject showing the distinct azeotropes studied with column. However, in azeotropic distillation, both the entrainer
different IL anions and cations used as separating agents. Table 2 and components must be vaporized. Moreover, large amounts of
lists the nomenclature used for all IL cations and anions mentioned entrainer are usually needed, leading to large energy consumption
herein. It is important to highlight the fact that most of the studies, as compared to extractive distillation, making the extractive option
approximately 75%, were performed using 1-alkyl-3-methylimida- more attractive [11]. Other criteria, e.g., corrosion, prices, sources,
zolium based ILs. The range of anions utilized is generally much etc. should also be taken into consideration and, from the strictly
greater, with special enphasis on [Cl], [PF6], [BF4], [OTf], economic viewpoint, the use of the entrainer with the highest rel-
[NTf2] and [EtSO4]. ative volatility will always translate into the lowest total annual
The use of ILs as entrainers in separation technology is promis- cost (TAC) [12].
ing [8], and several patents [9] already protected its application as Conventional extractive distillation typically results in low lev-
an entrainer in extractive or azeotropic distillation. els of product purity since the entrainer coming from the bottom of
the solvent recovery column contains impurities which influence
the separation effect. Moreover, extractive distillation consumes
1.3. Energy and economic costs much more energy than liquid–liquid extraction. It is thus advis-
able to combine extractive distillation with another separation
Distillation is the most widely used separation process in the process such as azeotropic distillation, liquid–liquid extraction
chemical industry. Typically, distillation columns and their support and so on [13]. It is also important to mention that new hybrid
facilities account for about one-third of the total capital cost and alternatives are gaining advantage over traditional processes. As
more than half of energy consumption. Consequently, the design an example, economic analysis showed that for the dehydration
6 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28

TABLE 2
Nomenclature of the ionic liquids used as separating agents.

Cations Anions
+ 
[MIM] N-methylimidazolium [Br] Bromide
[MMIM]+ 1-Methyl-3-methylimidazolium [Cl] Chloride
[EMIM]+ 1-Ethyl-3-methylimidazolium [PF6] Hexafluorophosphate
[PMIM]+ 1-Propyl-3-methylimidazolium [BF4] Tetrafluoroborate
[BMIM]+ 1-Butyl-3-methylimidazolium [OTf] Trifluoromethanesulfonate or Triflate or
Trifluoromethylsulfonate
[HMIM]+ 1-Hexyl-3-methylimidazolium
[OMIM]+ 1-Methyl-3-octylimidazolium [NTf2] Bis(trifluoromethylsulfonyl)imide
[DMIM]+ 1-Decyl-3-methylimidazolium [PF3(C2F5)3] Tris(pentafluoroethyl)trifluorophosphate
[DoMIM]+ 1-Dodecyl-3-methylimidazolium [OAc] Acetate
[EEIM]+ 1-Ethyl-3-ethylimidazolium [Salicylate] Salicylate
[AMIM]+ 1-Allyl-3-methylimidazolium [SCN] Thiocyanate
[EMMIM]+ 1-Ethyl-2,3-dimethylimidazolium [N(CN)2] Dicyanamide
[(EtOH)MIM]+ 1-(2-Hydroxyethyl)-3-methyl-imidazolium [C(CN)3] Tricyanomethanide
[(EtOH)MMIM]+ 1-(2-Hydroxyethyl)-2,3-dimethylimidazolium [B(CN)4] Tetracyanoborate
[(HeOH)MIM]+ 1-(6-Hydroxyhexyl)-3-methylimidazolium [BOB] Bis[oxalato(2-)]-borate
[Epy]+ 1-Ethylpyridinium [EtSO4] Ethyl sulfate
[EMpy]+ 1-Ethyl-3-methylpyridinium [MeSO4] Methyl sulfate
[3MBpy]+ 3-Methyl-N-butylpyridinium [OcSO4] Octyl sulfate
[4MBpy]+ 4-Methyl-N-butylpyridinium [Me(EtO)2SO4] Diethylenglycol monomethyl ether sulfate
[BMpyr]+ 1-Butyl-1-methylpyrrolidinium or 2-(2-methoxyethoxy)ethylsulfate
[P6 6 6 14]+ Trihexyl(tetradecyl)phosphonium [HSO4] Hydrogen sulfate
[EEES]+ Triethylsulfonium [MeSO3] Methanesulfonate
[MOOON]+ Methyltrioctylammonium [ToSO3] p-Toluenesulfonate
[(EtOH)NH3]+ Monoethanolammonium [(Bu)2PO4] Dibutylphosphate
[(EtOH)2NH2]+ Diethanolammonium [(Me)2PO4] Dimethylphosphate
[(EtOH)3NH]+ Triethanolammonium [(Et)2PO4] Diethylphosphate
[MMM(EtOH)N]+ (2-Hydroxyethyl)trimethylammonium [MePO3] Methylphosphonate
[EMM(EtOH)N]+ Ethyl(2-hydroxyethyl)dimethylammonium [(Me3Pe)2PO2] Phosphonium
[HMM(EtOH)N]+ Hexyl(2-hydroxyethyl)dimethylammonium Bis(2,4,4-trimethylpentyl)phosphinate
[EEM(MeOEt)N]+ Diethylmethyl(2-methoxyethyl)ammonium
Chloroaluminate ionic liquids
[BMIM][Cl]-AlCl3 1-Butyl-3-methylimidazolium chloride–aluminum
chloride
[MMMNH][Cl]- Trimethylamine hydrochloride–aluminum chloride
AlCl3
[EEENH][Cl]-AlCl3 Triethylamine hydrochloride–aluminum chloride

of isopropanol, a hybrid process consisting of distillation followed 1.4.1. Extractive or azeotropic distillation
by pervaporation with ceramic membrane, is the most economical Extractive distillation is a common technique for removing one
process leading to a total cost reduction of approximately 49% [14]. of the mixture components at or close to its azeotropic point. It in-
The use of ILs as entrainers in extractive distillation can over- volves the addition of a new heavy chemical compound (entrainer)
come the purity issues since, at the range of temperature used, which interacts with the components by altering their relative
the vapor pressure of the IL is still negligible. In extractive distilla- volatilities.
tion, the heat demand in the reboiler can be reduced, for example, Organic solvents, inorganic salts or even hyperbranched poly-
by lowering reflux ratios through the selection of favorable entra- mers are the most commonly used entrainers. Extractive distillation
iners. This optimization promises savings of at least 25% as com- with ILs as the separating agent integrates the advantages of a liquid
pared to conventional processes [8]. solvent (easy operation) and a solid salt (high separation ability).
Extractive distillation with ILs as the separating agent can Azeotropic distillation is also common in the separation of close
potentially result in high-purity products. ILs can overcome most boiling temperature or azeotropic mixtures; this technique requires
of the problems associated with the use of solid salts such as dis- the vaporization of both the entrainer and components into the top
solution, reuse, and transport. Other important advantages are: of an azeotropic distillation column. In this way, azeotropic distilla-
(1) no traces of entrainers in the distillate due to their nonvolatility tion is not used when the entrainer is an ionic liquid [13].
under the required conditions; (2) significant reduction of heat du- Another advantage of ILs is their negligible vapor pressure at
ties due to high selectivities and easy regeneration of the entrainer; ambient temperature and pressure conditions. This minimizes
and (3) the possibility of tailoring the properties of the IL. the chance of loss to the atmosphere, thus reducing environmental
In respect to liquid–liquid separation, Meindersma and de Haan and worker exposure issues; it further increases the ILs’ recyclabil-
[15] showed that, for the aromatic/aliphatic separation, the invest- ity and reusability, by stripping, evaporation or drying, thereby
ment (including extraction solvent) in a process using ILs is about minimizing costs. The relative volatility, aVLE, and selectivity, SVLE,
65% of that of the conventional sulfolane process and the heat load can be calculated from (vapor + liquid) equilibria (VLE) data and
is only about 20%. are defined as:
ðyi =xi Þ
aVLE
ij ¼ ; ð1Þ
1.4. Separation process ðyj =xj Þ
awith
ij
IL
The scope of the review is limited itself to discussing separation SVLE ¼ : ð2Þ
aij
without IL
processes utilizing ILs as separating agents.
 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28 7

1.4.2. Liquid–Liquid extraction 2. Azeotrope separation

Liquid–liquid extraction is based on the immiscibility between
two liquid phases and it is used to separate a dissolved component 2.1. Aqueous systems
(solute) from its diluent (inert) by its transfer into a second compo-
nent (extraction solvent) liquid phase. The equilibrium in extrac- 2.1.1. (Water + ethanol) system
tion can be characterized by the values of distribution coefficient, Due to its industrial importance and the availability of compo-
bLLE, and selectivity, SLLE. These parameters are widely used in nents, the (ethanol + water) system is by far one of the most stud-
assessing the solvent feasibility in the liquid–liquid extraction ied. Ethyl alcohol is produced by two main processes, acid
and are defined as follows: catalyzed hydration of ethylene and fermentation of sugars [18].
Hydration of ethylene is the primary method for the industrial pro-
xIL Phase
duction of ethyl alcohol, while fermentation produces most bever-
bLLE ¼ Solute
; ð3Þ
Organic Phase
xSolute age alcohol. Industrially produced ethanol has many uses: fuel and
! ! gasoline additive [19]; as a solvent in the manufacture of var-
xOrganic Phase
xIL Phase
nishes, lakers, inks, paints, and cleaning products; in the prepara-
SLLE ¼ Inert
 Solute
: ð4Þ
xIL Phase
Inert
Organic Phase
xSolute tion of essences and perfumes in the cosmetic industry [20]; as
flavorings and thickeners in the food industry; as an extraction sol-
The distribution coefficient determines the amount of IL required
vent and carrier in medicine and the pharmaceutical industry; and
for the extraction process and is thus related to the capacity of
finally as a powerful disinfectant, anti-microbial and anti-staling
the IL while the selectivity evaluates the efficiency of the IL used
compound [21]. All beverage alcohol and much of that used in
as solvent. Selectivity also indicates the ease of extraction of a sol-
industry are synthesized through fermentation of a variety of prod-
ute from the diluent (inert) and the distribution coefficient indi-
ucts including grains such as corn, potato mashes, fruit juices, and
cates the solute-carrying capacity of the solvent. Higher
beet and cane sugar molasses [22]. It, therefore, is an important
selectivity corresponds, theoretically, to fewer stages required for
bio-refinery product. One of the major drawbacks of switching to
a given separation and less residual inert in the extract. A higher
a biomassbased industry is the production cost of the base alco-
distribution coefficient corresponds to a lower solvent flowrate for
hols. The high affinity of ethanol for water, which makes it partic-
given separation, a smaller-diameter column, and lower operating
ularly useful in industrial applications, turns out to be a problem in
costs.
its purification. Purifying an alcohol requires about 6% of the ener-
The extraction efficiency indicates the ability of the solvent, IL,
getic value of the compound itself, much of which is used in sepa-
to remove solute from the azeotropes. This parameter can either be
rating and concentrating the alcohol from the fermentation broth
calculated from (liquid + liquid) equilibria (LLE) data, ELLE, or from
where it is produced [23].
extraction process data, EEP, and is defined as follows:
Separating alcohols from water conventionally requires an
energetically costly series of distillation columns. The (vapor +
C IL Phase IL Phase liquid) equilibrium of {ethanol + water} system has been subject
Solute  V Solute
ELLE ¼  100; ð5Þ of innumerable studies. In the Dortmund data bank there are
C Mix Mix
Solute  V Solute 127 references to 245 data sets reported for this system [24]. At
wFeed Raffinate
Inert  wInert atmospheric pressure, ethanol + water demonstrates a minimum
EEP ¼ Equilibrium
ð6Þ
wFeed
Inert  wInert : boiling temperature homogeneous azeotrope at xEthanol = 0.897
and T = 351.25 K (table 1). Many entrainers have been used.
Among them, inorganic salts have been successfully implemented.
1.4.3. Supported liquid membranes
Due to their polarity, inorganic salts can be easily dissolved in the
Membranes are functional materials for many separation pro-
mixture, creating a noticeable salt effect. In the database
cesses including bioseparations. Such processes are energy effi-
ELVḾ2003 there are 102 references to using 46 inorganic salts
cient, can be operated under moderate temperature and pressure,
to move or break the ethanol + water azeotrope [25]. The most
and do not produce large amounts of waste solvents. Supported li-
commonly used inorganic salts are calcium chloride and potas-
quid membranes (SLMs) are porous solids impregnated with the li-
sium acetate. Nevertheless, the use of salt distillation process
quid used for the separation. Conventional SLMs have two
may erode the sieve plate, precipitate and build up in the tower
common problems, e.g., the loss of liquid through vaporization or
due to the causticity and limited solubility [26]. In order to avoid
its dissolution in the contacting phases and the loading limits
these problems conventional entrainers such as N-methylpyrroli-
[16,17]. Since ILs are nonvolatile and are able to selectively modify
done (DMP), dimethylsuphoxide (DMSO) and sulpholane, are also
the properties of the membrane solvent, they have been advanta-
used.
geously incorporated into the porous membranes to form the so-
Table 3 lists research data (or calculated from them) regarding
called functional supported ionic liquid membranes (SILMs). SILMs
the ionic liquids used as azeotrope breakers for the (water + etha-
are versatile in a number of applications, including azeotropic mix-
nol) system for extractive distillation, liquid–liquid extraction and
tures separations.
supported liquid membrane [26–48].
To clarify the controlling factor in the permeation process, the
selectivity of the permeability, SMEM, is calculated in the following
equation: 2.1.1.1. Separation by extractive distillation. An analysis of the arti-
cles related to the use of entrainers for the (ethanol + water) binary
PermSolute system shows that most of the work has been done through the
SMEM ¼ : ð7Þ
PermInert evaluation of VLE behavior of ternary systems of (etha-
nol + water + entrainer) so that the feasibility of extractive or aze-
The separation factor for the vapor permeation, aMEM, is defined as
otropic distillation technique can be proposed.
follows:
Most of the ILs studied so far as entrainers for the (etha-
nol + water) system show a salting out effect, which enhances
ðpPermeate =pPermeate
Inert Þ
aMEM ¼ Solute
: ð8Þ the relative volatility of the ethanol and may even eliminate the
ðpSolute =pFeed
Feed
Inert Þ azeotropic phenomena at a specific IL content.
8 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28

TABLE 3
Ionic liquids used as azeotrope breakers on literature for the separating of water + ethanol, ethanol/water selectivity from LLE data (SLLE), ethanol/water relative volatility from
VLE data (aVLE) at azeotropic composition and ethanol/water separation factor from membranes data (aMEM).

Azeotrope Ionic liquids SLLE aVLE a aMEM Remarks Reference

Water + ethanol [EMIM][OTf] 1.27 VLE data at 100 kPa [27]
[EMpy][EtSO4] 2.40b VLE data at 101.3 kPa [28]
[BMIM][Cl] 1.38 VLE data at 30, 40 100 kPa [29]
[BMIM][OTf] 1.71c VLE data at 314.2 and 331.7 K [30]
[OMIM][OTf] 0.95d
[HMIM][Cl] 1.31 VLE data at 100 kPa [31]
[BMIM][MeSO4] 2.38b VLE data at 101.3 kPa [32]
[EMIM][EtSO4] 2.02b VLE data at 101.3 kPa [33]
[EMIM][BF4] 1.37 VLE data at 100 kPa [34]
[BMIM][BF4] 1.37
[EMIM][N(CN)2] 1.36
[BMIM][N(CN)2] 1.53
[EMIM][Cl] 1.68
[BMIM][Cl] 1.41
[EMIM][OAc] 1.63
[BMIM][OAc] 1.46
[HMIM][Cl] 0.39 VLE data at 101.3 kPa [35]
[BMIM][Cl] 1.87b VLE data at 101.3 kPa [36]
[MMIM][(Me)2PO4] 1.17e VLE data at 101.3 kPa [26]
[EMIM][(Et)2PO4] 1.41e
[BMIM][Br] 1.06e
[BMIM][Cl] 1.25e
[BMIM][PF6] 1.19e
[EMIM][BF4] 1.44 VLE data at 363.15 K [37]
[BMIM][BF4] 1.34
[BMIM][Cl] 1.55
[EEIM][(Et)2PO4] n.a. Vapor pressure data [38]
[EMIM][EtSO4] n.a. Vapor pressure data [39]
[EMIM][(Me)2PO4] n.a. Vapor pressure data [40]
[MMIM][(Me)2PO4] n.a. Vapor pressure data [41]
[MMIM][(Me)2PO4] n.a. Vapor pressure data [42]
[PMIM][Br] n.a. Vapor pressure data [43]
[(EtOH)NH3][OAc] n.a. Boiling temperature at 101.3 kPa [44]
[(EtOH)3NH][OAc] n.a.
[(EtOH)2NH2][Cl] n.a.
[P6 6 6 14][N(CN)2] n.a. LLE data at 298.15 K [45]
[P6 6 6 14][((Me)3Pe)2PO2] n.a.
[HMIM][ NTf2] 10f LLE data at 295 K [46]
[BMIM][PF6] 0.40g LLE data at 298.15 K [47]
[HMIM][PF6] 0.33h
[OMIM][PF6] 0.29i
[BMIM][PF6] 4–6 Supported liquid membranes: vapor permeation [48]
a
Ethanol molar fraction  0.95 and IL molar fraction  0.10.
b
Ethanol molar fraction  0.55 and IL molar fraction  0.10.
c
Ethanol molar fraction  0.90 and IL molar fraction  0.50.
d
Ethanol molar fraction  0.95 and IL molar fraction  0.80.
e
Ethanol molar fraction  0.95 and IL molar fraction  0.05.
f
Ethanol molar fraction in organic phase = 0.04.
g
Ethanol distribution coefficient at ethanol molar fraction = 0.97.
h
Ethanol distribution coefficient at ethanol molar fraction = 0.48.
i
Ethanol distribution coefficient at ethanol molar fraction = 0. 60.

2.1.1.1.1. Effect of the nature of the anion. The effect of the ionic li- in the mixture can be very valuable for understanding these low
quid’s anion nature on the relative volatility of the system at two relative volatilities at the azeotropic point. However, the analysis
ethanol concentrations was studied at xEthanol = 0.1 (supporting of the activity coefficients of several ILs with water and with etha-
information figure S1) and 0.95 (figure 2). Imidazolium cation nol [34] shows that the interaction of the anions with water and
was chosen due to the wealth of available data. There is a large dif- with ethanol follow the same order: [OAc]  [Cl] > [N(CN)2] >
ference in the relative volatility at or near the azotropic point and [BF4], canceling out most of the effect. This may explain the fact
at low ethanol concentrations. The separation factors are much lar- that no dramatic increase in the relative volatility has been
ger (around 10 for the most effective ILs) at low ethanol concentra- observed.
tions than at or near the critical point (around 2 for the most 2.1.1.1.2. Effect of the nature of the cation. Only two different cat-
effective ILs). At the azeotropic point, the most promising IL anions ions, [EMIM]+ and [EMpy]+ with the anion [EtSO4], have been
are [Cl] and [OAc]. The dialkyl phosphate based ILs also perform tested as entrainers (see supporting information, figure S2) for
very well at low concentrations of ethanol. Note that these anions the (ethanol + water) system [28,33]. At low ethanol concentra-
are also present in the most effective inorganic salts used for this tions, [EMIM][EtSO4] leads to higher relative volatility than [EM-
system [25]. py][EtSO4]; at xEthanol = 0.2, both ILs have similar effects on the
It is well known that the anion has a stronger role in interaction relative volatility of this system.
with a given solvent than the cation. Thus, the knowledge and eval- 2.1.1.1.3. Effect of the alkyl chain length. The length of the alkyl side
uation of the interactions between the anion and both components chain of both the anion and the cation is another important feature.
 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28 9

TABLE 4
Minimum molar fraction of IL needed to break the water + ethanol azeotrope at
xEthanol = 0.95. All the data are from reference [34], except those for [EMIM][OTf] [27].

Cation [Cl] [OAc] [N(CN)2] [BF4] [OTf]

[EMIM]+ 0.015 0.013 0.019 0.018 0.023
[BMIM]+ 0.019 0.015 0.019 0.023

concentrations. This indicates that small cations possess stronger

IL + water interactions than large ones, forcing the ethanol into
the vapor phase, thus increasing relative volatility. It is also worth
noting that the lack of any significant effect of the IL cation’s chain
length at low IL concentrations (xIL < 0.03).
An interesting study of the combined effect of the chain length
in both the anion and the cation (see supporting information, fig-
ure S3), using [MMIM][(Me)2PO4] and [EMIM][(Et)2PO4] [26] leads
FIGURE 2. Effect of the anion nature in ILs based on imidazolium cation [26,27,32– to the same conclusions about the alkyl chain length. For xEtha-
34] taking into account the relative volatility for the (ethanol + water) system at
nol > 0.40, both ILs perform similarly, with relative volatilities
xEthanol  0.95 and xIL  0.1.
approaching 1 at the azeotropic point; away from the azeotropic
point (xEthanol = 0.1) and xIL < 0.2, the smaller the IL the higher the
relative volatility.
2.1.1.1.4. Effect of the ionic liquid concentration. Another important
parameter to evaluate is the effect of IL concentration on the rela-
tive volatility of (ethanol + water) system. The following ILs were
studied: [EMIM][OTf] (xIL = 0.05, 0.1, and 0.2) [27]; [EMIM][BF4]
and [BMIM][BF4] (xIL = 0.1, 0.3, and 0.5) [37]; [MMIM][(Me)2PO4]
and [EMIM][(Et)2PO4] (wIL = 0.1 and 0.2) [26]; [BMIM][Cl] [29]
and [HMIM][Cl] [31] (xIL = 0.1, 0.2, 0.4, 0.6, 0.8, and 0.98). Behaviors
differed according to the IL. Figure 4 clearly shows that for
[HMIM][Cl] there is an inversion of behavior around xEthanol = 0.4;
at low ethanol concentrations there is salting in effect while salting
out occurs at high ethanol concentrations. The effect of [HMIM][Cl]
[31] on the phase behavior (figure S4) is also interesting. While the
boiling temperature of water is higher than that of ethanol, the
boiling temperature of [HMIM][Cl] + water is lower than that of
FIGURE 3. Effect of the IL cation chain length for chloride based ILs on the relative [HMIM][Cl] + ethanol at wIL = 0.8. Although at wIL = 0.1 the etha-
volatility for the (ethanol + water) system at xEthanol = 0.95. [EMIM][Cl] and
nol + water azeotrope is broken, it is clear that there is always a
[BMIM][Cl] data from [34]; [HMIM][Cl] from [31].
minimum for the bubble line and the dew line at a given wIL.
In conclusion, the effect of the IL concentration on the relative
volatility might not be as straightforward as expected since it is
‘‘slower’’ among different ILs. This is due to the balance of interac-
tions between the IL-water and IL-ethanol [31].
Table 4 lists the minimum amount of IL needed to break the
ethanol + water azeotrope for the several anions combined with
[EMIM]+ and [BMIM]+ cations. Very small concentrations of each
IL suffice. However, [EMIM][Cl] and [BMIM][Cl] have melting tem-
peratures about (358 and 340) K [50], respectively, and high vis-
cosities [51] and thus their use as an entrainers does not present
the usual IL advantages over inorganic salts. Furthermore, [BMI-
M][OAc] starts to decompose at 446 K and has a viscosity of
440 mPa s at 298.15 K. The viscosity data for [EMIM][OAc] is above
129 mPa s. The solution viscosities of these four ILs with water and
ethanol are close to the acceptable distillation limit of 100 mPa s,
above which it becomes very difficult to manage the separation.
FIGURE 4. y versus x representation for (ethanol + water) system at variable
[EMIM][OTf] has a melting temperature of 262.23 K [52], a viscos-
concentrations of [HMIM][Cl] [31]. Ethanol + water data from [49]. ity of 41 mPa s [53] and a decomposition temperature higher than
others, 613 K. Its only drawback is its high production cost.
2.1.1.1.5. Comparison with other entrainers. Figure 5 demonstrates
The cation chain length effect ([EMIM]+ and [BMIM]+) was investi- an evaluation of the performance of several entrainers proposed
gated for the following anions: [OAc], [N(CN)2], [BF4] [34]. For for the (ethanol + water) system. The two polymers (hyper-
the [Cl] anion studies were performed using [EMIM]+ [34], branched poly(ethylene glycol) [54] and polyesteramide Hybrane
[BMIM]+ [26,29,34,36,37], and [HMIM]+ [31,35]. Figure 3 repre- S1200 [55]) and the inorganic salt (CaCl2) [56] were chosen on
sents studies conducted near the azeotropic point [31,34]. For all the basis of the best separation factors for that particular family
these ILs, the increase in the cation’s chain length led to a less of entrainers. 1,2-Ethanediol [8] is the conventional entrainer and
effective entrainer, since the relative volatility decreases at all [BMIM][Cl] is one of the most promising novel candidates.
10 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28

FIGURE 5. Comparison of the relative volatility for [BMIM][Cl] [29] and other
proposed entrainers: hyperbranched polyglycerol (PG1) [54]; hybrane S1200 FIGURE 6. Experimental vapor pressure data for the systems (ethanol + water)
(S1200) [55]; calcium chloride (CaCl2) [56]; 1,2-ethanediol (ETH) [8]. (dashed-point line) [57] and (ethanol + water + IL) [41,38,40] at xEthanol  0.9
(azeotropic point) and xIL  0.05.

CaCl2 is the most overall effective entrainer with higher separa-

Recent development in membrane technology has improved the
tion factors than all the other compounds, especially at low ethanol
efficiency of the (ethanol + water) system separation. [BMIM][PF6]
concentrations. However, as mentioned before, when solid salt is
was supported on a porous membrane in a vapor permeation appa-
used in industrial operations, dissolution, reuse and transport of
ratus [48] with a resulting separation factor of 4 to 6 during 100 h.
salt present considerable challenges, not to mention the concur-
These results are very encouraging since the rapid fouling of the
rent jam and erosion. Due to these issues, IL entrainers represent
membranes is a major issue with this technique.
a step forward in the use of extractive distillation to break etha-
nol–water azeotrope. In addition, at the azeotropic point, both
[BMIM][Cl] and CaCl2 have very similar separation factors. 2.1.2. (Water + other alcohols) systems
This section deals with the azeotropic separation of mixtures of
longer (alcohols + water). The alcohols already studied in the liter-
2.1.1.2. Separation by other methods. Other methods to identify and
ature include 1-propanol, 2-propanol, 2-methyl-2-propanol and n-
evaluate the suitability of ILs as entrainers were vapor pressure
butanol. 1-Propanol is mainly used as a solvent, namely as a carrier
measurements (6 articles), boiling temperature (1 article), (liqui-
and extraction solvent for natural products. Due to its solvent and
d + liquid) equilibria (3 articles) and the use of supported ILs (1
antiseptic properties, it is also used in drugs and cosmetics [58]. 2-
article). Although vapor pressure and boiling temperature mea-
Propanol is also used in pharmaceutical, household, and personal
surements are in fact VLE experiments, they were considered sep-
products [59]. 2-Methyl-2-propanol is used for the removal of
arately since the results lead to conclusions which differ from
water from substances, in the extraction of drugs, in the manufac-
those of the VLE data.
ture of perfumes, in the recrystallization of chemicals, and as a
As in the case of VLE experiments, the entrainer interacts with
chemical intermediate [60]. Butanol is an important bulk chemical
both solvents producing, occasionally, unexpected results. The
with a wide range of industrial uses e.g., surface coatings, superab-
addition of phosphate based 1,2 dialkylimidazolium ILs in the va-
sorbents, fibers, plastics, resins, hard-surface cleaners, electronics,
por pressure of the (ethanol + water) system was tested [41,42].
paints, as synthetic fruit flavoring product, pharmaceuticals and
The obtained results are identical to those obtained when VLE
perfume manufacturing [61]. Mixtures of these longer alcohols
measurements were performed [26]; the smaller the alkyl substi-
with water occur frequently. Therefore, their recovery from aque-
tuting hydrocarbon chain of the IL proves more effective. Binary
ous solutions is a topic of great interest.
data for ethanol and a series of phosphate-based ILs indicate that
The study of azeotrope breaking using ILs in binary systems of
the ILs can be ordered as is [MMIM][(Me)2PO4] > [EMIM][(Et)2-
longer alcohols and water is just beginning. Up to now, only 10 ref-
PO4] > [BMIM][(Bu)2PO4]. Thus only the vapor pressure of etha-
erences of VLE [62–71], four of LLE [46,72–74], and one of mem-
nol + water + [MMIM][(Me)2PO4] was measured. Figure 6 plots
brane separation [48], have been found. In terms of ternary
the tests of [EMIM][(Me)2PO4] [40] and [EEIM][(Et)2PO4] [38]. It
azeotropic mixtures, there is only one paper [75] dealing with
can be observed that, close to the azeotropic point, both ILs per-
(water + ethanol + ethyl acetate) system. Table 5 lists the ILs used
form equally well. The results from the first IL, [MMIM][(Me)2PO4],
as azeotrope breakers for the (water + alcohols) longer than etha-
[41,42] are similar to those obtained by VLE [26] with the salting-
nol binary azeotropes and for the (water + ethanol + ethyl acetate)
out of ethanol at all ethanol concentrations. However, for the sec-
ternary azeotrope [62–75].
ond IL, salting-in occurs at low ethanol concentration while salting
out is obtained near the azeotropic point. For [MMIM][(Me)2PO4],
this salting-in/salting-out change with the ethanol concentration 2.1.2.1. Separation by extractive distillation. The results available
also takes place. Vapor pressure data of [EMIM][EtSO4] [39] and highlight the feasibility of using ILs as entrainers for the separation
[PMIM][Br] [43] corroborate the trends for the VLE data. of (water + alcohol) systems. As already shown for the
Only three articles address the ethanol + water azeotrope using (water + ethanol) system, most of the ILs showed a salting out
LLE. Chowdhury et al. [45] used long chain phosphonium based ILs. effect, enhancing the relative volatility of the alcohol and even
Chapeaux et al. [46] and Swatloski et al. [47] used [HMIM]+ based eliminating the azeotropic phenomena at specific IL contents.
ILs, with the [NTf2] and [PF6] anions respectively. For low con- 2.1.2.1.1. Binary azeotropic mixtures. Effect of the nature of the
centrations of ethanol in water (typically xEthanol = 0.04 in a fer- anion and cation. The long list of commonly used anions with dis-
mentation broth) the ethanol distribution coefficients are around tinctive properties, requires the study of the anion effects on the
10. The authors conclude that large amounts of this IL are required chosen separation process. The study of azeotrope breaking of
to separate diluted ethanol aqueous solutions. the system (water + 1-propanol) has only been carried out with
 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28 11

TABLE 5
Azeotropes in the literature regarding the separation of water + longer alcohols. ILs used as azeotrope breakers. 1-Butanol/water selectivity from LLE data (SLLE). Alcohol/water
relative volatility from VLE data (aVLE) and 1-propanol/water separation factor from membranes data (aMEM).

Azeotrope Ionic liquids SLLE a

aVLE aMEM Remarks Reference
b
Water + 1-propanol [EMIM][OTf] 1.36 VLE data at 100 kPa [62]
[EMIM][BF4] 1.52c VLE data at 100 kPa [63]
[BMIM][BF4] 1.37c
[BMIM][PF6] 6–8 Supported Liquid Membranes: Vapor Permeation [48]
Water + 2-propanol [EMIM][EtSO4] 1.41d VLE data at 333.15 K: headspace gas chromatography [64]
[BMIM][BF4] 1.51d
[HMIM][NTf2] 1.41e VLE data at 332 K: headspace gas chromatography [65]
[BMpyr][NTf2] 1.51e
[BMIM][BF4] 1.32d VLE data at 101.3 kPa [66]
[EMIM][BF4] 1.95d VLE data at 101.3 kPa [67]
[BMIM][BF4] 1.32d VLE data at 100 kPa [68]
[EMIM][BF4] 2.19d VLE data at 100 kPa [63]
[BMIM][OAc] 2.12b VLE data at 100 kPa [69]
[EMIM][OAc] 2.18b
[BMIM][Cl] 2.20b
[EMIM][N(CN)2] 1.74b
[EMIM][BF4] 1.58b
[BMIM][N(CN)2] 1.50b
[BMIM][BF4] 1.38b
[EMIM][NTf2] n.a. Vapor pressure data [70]
[BMIM][NTf2] n.a.
Water + 2-methyl-2-propanol [EMIM][OAc] 3.28b VLE data at 100 kPa [71]
[BMIM][OAc] 2.58b
[HMIM][OAc] 2.43b
[EMIM][Cl] 3.58b
[BMIM][Cl] 2.29b
[HMIM][Cl] 2.05b
Water + 1-butanol [(HeOH)MIM][NTf2] 40 LLE data at 295 K [72]
[HMIM][PF3(C2F5)2] 300
[HMIM][NTf2] 90 LLE data at 295 K [46]
[(EtOH)MMIM][BF4] n.a. LLE data at 293 K [73]
[(EtOH)MIM][BF4] n.a.
[BMIM][PF6] 21 LLE data at 296 K [74]
[OMIM][PF6] 37
Water + ethanol + ethyl acetate [BMIM][OAc] 14/84.1/11.3f VLE data at 100 kPa [75]
a
1-Butanol mass fraction in the initial aqueous solution  0.05.
b
2-Propanol molar fraction  0.95 and IL molar fraction  0.25.
c
1-Propanol molar fraction  0.43 and IL molar fraction  0.24.
d
2-Propanol molar fraction  0.67 and IL molar fraction  0.20.
e
Water/2-propanol relative volatility; 2-propanol molar fraction  0.38 and IL molar fraction  0.95.
f
Water molar fraction  0.04 and IL mass fraction  0.7 varying the (ethyl acetate molar fraction/(ethyl acetate molar fraction + ethanol molar fraction)) ratio. Results
indicated are for relative volatility of ethyl acetate to ethanol/ethyl acetate to water/ethanol to water, respectively.

two distinct anions. The minimum molar fraction of IL needed to

break the water + 1-propanol azeotrope at x1-Propanol  0.43 is
smaller for [EMIM][BF4], 0.30 [63], than for [EMIM][OTf], 0.34
[62]. The small relative volatility (see table 5) obtained for the
[OTf] anion IL (compared to that of the [BF4]) demonstrates its
inferior entrainer capacity.
Results for the system (water + 2-propanol) cover a much larger
spectrum of ILs. Three distinct experimental working conditions
were selected for comparison, one at x2-Propanol  0.95, another
close to the azeotrope composition and the last one at xIL = 0.95.
VLE data, namely with [EMIM][BF4], [BMIM][BF4], [EMIM][N(CN)2],
[BMIM][N(CN)2], [EMIM][OAc], [BMIM][OAc], or [BMIM][Cl], were
taken at x2-Propanol  0.95 [69] and are compared in figure 7. The re-
sults indicate that the effect on the relative volatility of 2-propanol
to water is mainly dependent on the anion in the order: [Cl] >
[OAc] > [N(CN)2] > [BF4]. This order is in agreement with that
found for water + ethanol azeotrope.
Only Kim et al. [64], have performed studies with a distinct an- FIGURE 7. Effect of the IL [69] on the relative volatility for the (water + 2-propanol)
ion worth mentioning. Besides using [BMIM][BF4], the authors system at x2-Propanol  0.95.

have also used [EMIM][EtSO4] as an azeotrope breaker for

water + 2-propanol. At the concentrations used, the authors were
not able to break the azeotrope, rather deviating it to higher con- a mol fraction of 0.2 of [EMIM][EtSO4] and [BMIM][BF4], respec-
centrations of alcohol, from xAlcohol = 0.68 to 0.88 and 0.84 using tively. Direct comparison of the relative volatility (table 5) for
12 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28

FIGURE 8. Effect of the alcohol size on the alcohol/water selectivity using distinct ILs [34,69,71] as azeotrope breaker at xAlcohol = 0.95.

[BMIM][BF4] and [EMIM][EtSO4] shows that the [EtSO4] anion is organic entrainers is available in patent literature [76,77]. Both
not a good option as an azeotrope breaker for this system. patents study distinct alternatives and give examples of the best
The only report [71] about the azeotrope (water + 2-methyl-2- entrainers for extractive distillation. The experimental conditions
propanol) evaluates the effect of six distinct ILs (supporting in example 1 of Berg and Yang [76] (azeotrope 2-propanol + water
information figure S5) based on the anions [Cl] and [OAc] and with a molar fraction of the conventional entrainer (propyl buty-
the cations [EMIM]+, [BMIM]+, [HMIM]+. The authors used a rate) of approximately 0.15) are close to those shown in table 1
constant molar fraction of alcohol of 0.95 and the anion ordering for 2-propanol. The relative volatility with propyl butyrate, 1.72,
is in close agreement with that for 2-propanol and ethanol. is lower than the average, 2.1, with [EMIM][BF4] [63,67], which
The correct choice of the IL cation may also improve the separa- was not the best IL entrainer. In the case of the azeotrope 1-propa-
tion process. However, regarding VLE data, Westerholt et al. [65] nol + water, example 3 of Berg and Yang [76], the experimental
has been the only study to address the influence of the nature of conditions are also similar to those used with ILs as azeotrope
the cation on the behavior of the azeotrope (water + 2-propanol) breakers. In this case, the relative volatility with the conventional
(supporting information figure S6). [HMIM][NTf2] and entrainer (butoxypropanol) is 2.2, which is higher than that ob-
[BMpyr][NTf2] served as entrainers. It is important to stress the ex- tained with [EMIM][BF4]. This proves that ILs can perform better
tremely high IL concentration used in this study, xIL  0.85 to 0.95. than conventional entrainers.
Since [NTf2] based ILs are usually hydrophobic, the objective was The addition of inorganic salts also breaks the azeotropes
to solubilize the alcohol and evaporate the water. The results water + 1-propanol and water + 2-propanol by producing an
clearly show that the proposed ILs, at the concentrations used, important salting out effect over the alcohol [78–82]. In conclu-
are not an alternative for this separation. sion, inorganic salts are more effective than ILs, strictly from the
Effect of the alkyl chain length. Only one study of the impact azeotrope breaking point of view. Nevertheless, ionic liquids offer
of increasing the alkyl chain length of the cation in the azeotrope clear operational advantages.
water + 1-propanol is available [63]. Two distinct 1-alkyl-3-meth- Effect of the size of the alcohol. The experimental data high-
ylimidazolium based cations with [BF4] anion were used: lights the effect of the alcohol hydrocarbon chain size in the sepa-
[EMIM]+ and [BMIM]+. The results indicate that, at the same weight ration of the azeotropes water + alcohols. To normalize the data,
fraction of IL, wIL = 0.7, the azeotrope is broken only in the case of the relative volatility of each system was divided by the relative
the IL with [EMIM]+ cation. Thus, for this system, smaller alkyl volatility of the binary system without entrainer at the same molar
chain length of the cation is more adequate for azeotrope breaking. fraction obtaining the alcohol/water selectivity of each system. The
Figure 7 illustrates the effect of the alkyl chain length of the cat- binary systems data were obtained from the literature
ion on the relative volatility of 2-propanol to water, namely the or- [49,62,63,83]. Figure 8 shows that the alcohol/water selectivities
der [EMIM]+ > [BMIM]+. This is evident for all cation/anion pairs are close to unity at low IL mass fraction, but increase with increas-
studied. The IL concentration effects on the relative volatility also ing IL concentration. For all ILs, the following order is obtained: 2-
appear. Several other articles studied [BF4] based ILs with both methyl-2-propanol > 2-propanol > ethanol.
[EMIM]+ and [BMIM]+ as cations as azeotrope breakers for the sys- 2.1.2.1.2. Ternary azeotropic mixture. Zhang et al. [75] analyzed the
tem (water + 2-propanol) [63,66–68]. All results confirm this trend, effect of [BMIM][OAc] on the ternary system (water + etha-
which is analogous to that found for the system (water + 1- nol + ethyl acetate) at low water mole fraction. The experiments
propanol). were performed at constant ethyl acetate/(ethanol + ethylacetate)
Comparison with other Entrainers. Data for the separation of ratio in the whole IL concentration range. As an example, at
propyl alcohols by extractive distillation using conventional wIL = 0.7 and xWater = 0.04, the relative volatilities ethyl acetate to
 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28 13

TABLE 6
Ionic liquids used as azeotrope breakers in the literature for the separating of water + tetrahydrofuran, water/THF selectivity from LLE data (SLLE) and THF/water relative volatility
from VLE data (aVLE).

Azeotrope Ionic Liquids SLLE a

aVLE b Remarks Reference
Water + Tetrahydrofuran [EMIM][BF4] 15.58 4.40 VLE data and LLE data at 337.15 K: headspace gas chromatography [37]
[BMIM][BF4] 3.79 3.24
[OMIM][BF4] 2.50
[BMIM][Cl] 37.26
[EMIM][Cl] 31.14 VLE data at 337.15 K: headspace gas chromatography [89]
[EMIM][EtSO4] 13.59
[EMIM][ToSO3] 11.41
[EMIM][OcSO4] 12.87
[EMIM][OTf] 3.92
[EMIM][BF4] 4.35
[EMIM][PF6] 2.21
[EMIM][NTf2] 1.50
[BMIM][OAc] 36.86
[BMIM][Cl] 31.35
[BMIM][HSO4] 8.50
[BMIM][Salicylate] 8.33
[BMIM][SCN] 8.68
[BMIM][ToSO3] 9.45
[BMIM][N(CN)2] 8.33
[BMIM][MeSO4] 8.62
[BMIM][Me(EtO)2SO4] 9.32
[BMIM][OcSO4] 22.03
[BMIM][BF4] 3.52
[BMIM][BOB] 15.67
[BMIM][NTf2] 1.26
[(EtOH)MIM][BF4] 14.79 LLE data at 293.15 K [90]
[(EtOH)MIM][Cl] 146.8
a
Water molar fraction in organic phase  0.05.
b
Azeotropic composition and IL molar fraction  0.30.

ethanol, ethyl acetate to water and ethanol to water can reach 14, This is a cyclic ether with manifold applications, for example as a
84.1 and 11.3, respectively at distinct ethylacetate/(ethanol + eth- solvent in the manufacture of paint, adhesives, pharmaceutical,
ylacetate) ratio. These results suggest the effectiveness of ILs as and other products, as an intermediate product and as a monomer
azeotrope breakers for ternary azeotropic systems. [84]. Its most important industrial application includes the produc-
tion of politetramethylene-glycol, which is mandatory in the man-
2.1.2.2. Separation by other methods. As in the case of etha- ufacture of polyurethane, polyester ether and Spandex fibers [85].
nol + water, vapor pressure measurements were considered sepa- The THF market has grown approximately 5.6% annually [86] in re-
rately since the information obtained does not allow for the cent years. It is produced by dehydrocyclization of 1,4-butanediol
same kind conclusions as VLE data. Doker and Gmehling [70] were in the presence of an acid catalyst. The produced (THF + water)
the first to address the problem of azeotrope breaking of the mixture presents an azeotrope at xTHF = 0.817 (table 1), and must
water + 2-propanol at 353.15 K using different concentrations of be dehydrated to xTHF = 0.999 [87] for commercial use. The key step
[EMIM][NTf2] and [BMIM][NTf2] and these ternary systems show in purification is the breaking of the THF + water azeotrope [88],
a positive deviation from Raoult’s law. While the 2-propanol is which is generally achieved either by pressure-swing (low/high
miscible with the two ILs over the entire composition range, the pressure), extractive, or azeotropic distillation.
two (IL + water) systems show an immiscibility gap beginning at Table 6 [37,89,90] lists the ionic liquids used as azeotrope
a water mole fraction of approximately 0.5. breakers for the (THF + water) system in the literature for extrac-
There is only one study [48] about separation of the azeotropes tive distillation and liquid–liquid extraction.
(water + alcohols) implementing supported liquid membranes.
[BMIM][PF6] was supported on the membranes and vapor perme- 2.1.3.1. Separation by extractive distillation. Most of the research
ation process was employed for the dehydration of aqueous 1-pro- regarding the use of ILs as entrainers for the (THF + water) binary
panol. A separation factor of 6 to 8 during 270 h occurred. These system studies the VLE behavior of ternary systems of (THF + wa-
results show a higher separation factor and experiment time than ter + IL) in order to evaluate the feasibility of extractive distillation
those mentioned previously for the water + ethanol. Supported li- techniques. Both the influence of systematic cation and the anion
quid membranes seem to be a promising method for azeotropic variation of imidazolium-based ILs with respect to their entrainer
separations. efficiency in extractive distillation have been investigated.
The study of ILs as solvents for liquid–liquid extraction is scarce All ILs studied (from table 6) are suitable as entrainers for sep-
and devoted only to the (water + 1-butanol) system. Evaluating the arating the azeotropic (THF + water) system by extractive distilla-
distinct ILs that were tested for the azeotropic separation by LLE of tion. Adding the selected IL to a binary (THF + water) system
1-butanol and water (1-butanol/water selectivities in table 5), it is leads to a remarkable increase in the molar vapor fraction of
clear that the best option is the [HMIM][PF3(C2F5)3] which provides THF, e.g., salting out effect, and thus the azeotrope is broken.
a higher distribution coefficient and selectivity. 2.1.3.1.1. Effect of the alkyl chain length. The effect of the variation
of the cation’s alkyl chain length on the VLE of the azeotropic sys-
2.1.3. (Water + tetrahydrofuran) system tem (THF + water) is depicted in figure 9. The relative volatility of
Tetrahydrofuran (THF) dehydration is a process of special eco- the low-boiling compound (THF) increases as the alkyl group is
nomical concern, due to increasing demand for anhydrous THF. shortened from an octyl group over a butyl group to an ethyl
14 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28

FIGURE 9. Influence of the variation of the length of the cation’s alkyl chain on the
FIGURE 11. Effect of the IL anion’s nature on the relative volatility of the
VLE of the azeotropic system THF + water. Experimental VLE data of the systems
(THF + water) azeotropic system at xTHF = 0.817 (azeotropic point) at 337.15 K
(THF + water + [BF4]) based ILs [37] at xIL = 0.5 and 337.15 K; binary THF + water
[89] and xIL = 0.30; binary THF + water data at 337.15 K [91]. The ILs cation is
data [91] at 337.15 K.
[BMIM]+.

group. If we accept that the IL entrainer suitability in extractive

systematic anion variation of specific cation-based ILs with respect
distillation rises with the increase in the relative volatility of the
to their entrainer efficiency is an important first step towards the
low-boiling component (with constant IL concentration) we can
task-specific tailoring of ILs for extractive distillation.
clearly see the entrainer superiority of [EMIM][BF4] as compared
The effect of the ionic liquid’s anion nature on the relative vol-
to corresponding imidazolium salts with longer alkyl chains.
atility of the (THF + water) system at the azeotropic point
Figure 10 depicts the effect of the IL concentration on the rela-
(xTHF = 0.817) is plotted in figure 11 using the [BMIM]+ cation (fig-
tive volatility of the (THF + water) system at azeotropic point
ure S7 illustrates the results for the [EMIM]+ cation). The relative
(xTHF = 0.817). An increase in IL concentration leads to a higher
volatility for the binary (THF + water) azeotropic system also ap-
THF concentration in the vapor phase and, therefore, to greater rel-
pears. All ILs in figure 11 are suitable entrainers for extractive dis-
ative volatility due to strong selective interactions between the IL
tillation of azeotropic (THF + water) systems, since the addition of
and water molecules. Figure 10 clearly shows that [EMIM][BF4] is
each IL to a binary (THF + water) system leads to a increase in the
the most suitable entrainer for the THF + water separation.
molar vapor fraction of THF. [BMIM][OAc], [BMIM][Cl], and [EMI-
Although it is well known that [EMIM][BF4] is chemically unsta-
M][OAc] exhibit the highest entrainer efficiencies.
ble within aqueous media, Jork et al. [37] argued that this IL is a
The relative volatilities of the THF + water + [BMIM][OAc],
suitable entrainer for long-thermal applications in thermal separa-
[BMIM][Cl], and [EMIM][Cl] (figure 11 and supporting information
tion. Despite its transition from slightly yellow to black, the ternary
figure S7) as compared to those of THF + water + [BMIM][BF4],
THF + water + [EMIM][BF4] VLE phase diagram was not found to
[EMIM][BF4], and [OMIM][BF4] (figure 10) obviously demonstrate
change at all. This statement does not warrant the chemical stabil-
that the exchange of the anion results in a much larger change in
ity of the IL investigated in a long-term application within aqueous
the relative volatility of the low-boiling component than the mod-
media. Nevertheless, it was suggested that chemical reactions
ification of the cation’s alkyl chain length. [BMIM][OAc] is shown
occurring in this mixture present minor impact on the observed
to be the most suitable for the THF + water separation. Moreover,
thermodynamic behavior.
the imidazolium salts containing the acetate ([OAc]) and chloride
2.1.3.1.2. Effect of the nature of the anion. Effective extractive distil-
lation entrainers must interact differently with the components of
the system to be separated. Thus, the study of the influence of

FIGURE 12. Effect of the chloride salt nature, inorganic versus IL, on the relative
volatility of the (THF + water) azeotropic system at xTHF = 0.817 (azeotropic point).
FIGURE 10. Influence of the IL cation chain length for [BF4] based ILs [37] on the The relative volatilities were determined experimentally at 337.15 K. VLE data of
relative volatility of the azeotropic system (THF + water) at xTHF = 0.817 (azeotropic the systems (THF + water + inorganic salts) saturated with each inorganic salt [91];
point). Experimental binary THF + water data from [91]. IL was always used at xIL = 0.30 [89].
 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28 15

FIGURE 13. Effect of ILs and hyperbranched polymers on the relative volatility of
the (THF + water) azeotropic system at xTHF = 0.817 (azeotropic point). The relative
volatilities with ILs (mole fraction of 0.30) were determined experimentally at
337.15 K [89]. The straight horizontal line represents the relative volatility with
hyperbranched polyesteramide Hydrane S1200 [92] (mole fraction of 0.40),
determined at 343.15 K.
FIGURE 14. Selectivity versus extraction solvents used in the separation of the
water + tetrahydrofuran azeotrope at solute molar fraction in organic phase  0.05
(for the ILs [37,90] solute = water and for the polymers [54,55]
([Cl]) anions are the best of the studied anions regarding the solute = tetrahydrofuran).
selective interactions and the increasing relative volatility of the
THF. This behavior is in accordance with the highest nucleophilic-
ity (e.g., ability to hydrogen bond) of the [OAc] and [Cl] anions. leads to different possible process options. The potential of ILs as
The imidazolium salts containing the [OAc] and [Cl] anions extractive distillation entrainers for the (THF + water) azeotropic
selectively interact with water. The formation of hydrogen bonds system has already been demonstrated. The aptness of ILs as
between these ILs and water molecules reduces water activity extraction solvents for the (THF + water) separation will now be
and thus increases relative THF volatility. addressed.
Hu et al. studied the use of Ils as extraction solvent for the azeo-
2.1.3.1.3. Comparison with other entrainers. In order to fully address
trope water + THF [90]. Jork et al. [37] used ILs in the extractive dis-
the entrainer efficiency of ILs in breaking the THF + water azeo-
tillation and solvent extraction for the separation of azeotrope.
trope, the performance of other entrainers such as inorganic chlo-
Figure 14 compares water/THF selectivity values for all the ILs
ride salts [91] and non-volatile hyperbranched polymers including
tested as solvents in liquid–liquid extraction. The use of [(EtOH)-
hyperbranched polyesteramide Hydrane S1200 [92] must be
MIM][Cl] as solvent leads to the highest selectivity values. The
compared.
same figure compares ILs and polymers [54,55] used as extraction
The pioneering work of Sada et al. [91], containing VLE data,
solvents. [(EtOH)MIM][Cl] continues to display the best results,
analyzes the effect of adding inorganic chlorides salts on the (va-
although the hyperbranched polyester Boltorn H3200 presents
por + liquid) equilibria of the (THF + water) azeotropic system at
better selectivity than other ILs.
fixed solvent composition and atmospheric pressure. Figure 12
The results, interestingly, also suggest the feasibility of specific
compares chloride inorganic salt versus IL regarding their effects
tailoring of the extraction solvent. Figure 15 plots the contrast be-
on the relative volatility of the (THF + water) system at azeotropic
tween IL and polymer behavior. While ILs are miscible in the water
point (xTHF = 0.817). The relative volatility for THF in the binary
and immiscible in THF, the polymers are miscible in THF and
(THF + water) azeotropic system is also shown. It can be concluded
immiscible in water. When the desired product is the THF, ILs
that the chloride ILs ([EMIM][Cl] and [BMIM][Cl]) perform better
may be used as solvents extracting water from azeotrope and
that the chloride inorganic salts (NaCl and CaCl2) in the salting-
obtaining THF in the upper phase. Otherwise, polymers are the best
out of THF from the aqueous solution (supporting information fig-
option.
ure S8), e.g., on increasing the relative volatility of the THF.
Some authors have recently suggested the use of nonvolatile
hyperbranched polymers as entrainers for extractive distillation
2.2. Alcoholic systems
[54,55,93,94]. Therefore, a comparison between the influence of
hyperbranched polymers and ILs on the VLE of (THF + water) aze-
2.2.1. (Alcohols + ester) systems
otropic system is required to access which additive performs better
In many chemical processes, (alkanol + ester) systems must be
in breaking the azeotrope. Figure 13 compares the effect of all
separated to obtain pure alcohol and pure ester. For example, these
investigated ILs, at a constant mole fraction of 0.30, on the relative
compounds constitute the products/reactants of further esterifica-
volatility of THF in the (THF + water) system at azeotropic point
tion reactions (esterification of methanol, ethanol, 1-propanol, and
(xTHF = 0.817) to that obtained for hyperbranched polyesteramide
1-butanol with acetic acid to obtain methyl acetate, ethyl acetate,
Hydrane S1200. The relative volatility for THF in the system
1-propyl acetate, and 1-butyl acetate). Azeotrope 2-propa-
(THF + water + hyperbranched polyesteramide Hydrane S1200)
nol + ethyl acetate, for example, is important for the solvent
(mole fraction of 0.40) is 1.87 [92], and is represented as a straight
extraction of edible oils [95]. The purification techniques of esters
horizontal line. It can be concluded that the hyperbranched poly-
such as azeotropic distillation [96], extractive distillation [97,98],
mer only performed better than two ILs, [EMIM][NTf2] and
and membrane separation [99] have been investigated. Table 7
[BMIM][NTf2].
(data from [100–108]) reports on ILs studied as azeotrope breakers
for separating (alkanols + ester) systems.
2.1.3.2. Separation by liquid–liquid extraction. An analysis of the lit- Ochilles et al. [100,103] studied extractive distillation for
erature regarding the use of ILs to break the THF + water azeotrope the separation of azeotropes methanol + methyl acetate and
16 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28

(a) (b)

FIGURE 15. Experimental tie-lines in mass fraction of the ternary systems: (a) (tetrahydrofuran + water + [(EtOH)MIM][Cl]) at 293.15 K [90] and (b) (tetrahydrofu-
ran + water + Hyperbranched polyester Boltorn H3200) at 334.15 K [55].

TABLE 7
Azeotropes in the literature regarding the separating of alcohols + esters. ILs used as azeotrope breakers. Alcohol/ester selectivity from LLE data (SLLE) and acetate/alcohol relative
volatility from VLE data (aVLE).

Azeotrope Ionic liquids SLLE a

aVLE b Remarks Reference
Methanol + methyl acetate [EMIM][OTf] 1.85 VLE data at 100 kPa [100]
Ethanol + ethyl acetate [EMIM][BF4] 1.83 VLE data at 101.32 kPa [101]
[BMIM][BF4] 1.45 VLE data at 101.32 kPa [102]
[OMIM][BF4] 2.35
[EMIM][OTf] 1.57 VLE data at 100 kPa [103]
[BMIM][Cl] 89c LLE data at 298.15 K [104]
[HMIM][Cl] 75c
[OMIM][Cl] 97c
[AMIM][Cl] 84c
[AMIM][Br] 81c
[EMIM][BF4] 1.56 LLE data at 298.15 K [105]
[(EtOH)MIM][BF4] 2.51
[EMMIM][BF4] 1.54
[(EtOH)MMIM][BF4] 1.93
[MIM][HSO4] 30.75 LLE data at 313.15 K [106]
[EMIM][HSO4] 19.61
[BMIM][HSO4] 10.26
1-Propanol + 1-propyl acetate [EMIM][HSO4] 21.51 LLE data at 313.15 K [107]
2-Propanol + ethyl acetate [BMIM][PF6] 3.29 LLE data at 298.15 K [108]
[HMIM][PF6] 2.65
[MMIM][MeSO4] 6.70
1-Butanol + 1-butyl acetate [EMIM][HSO4] 17.25 LLE data at 313.15 K [107]
a
Alcohol molar fraction in organic phase  0.10.
b
Azeotropic composition and IL molar fraction  0.30.
c
Extraction efficiencies from equation (5).

ethanol + ethyl acetate using [EMIM][OTf] as entrainer. This IL low-energy consumption. An analysis of the alcohol/ester selectiv-
showed a notable salting-out effect, enhancing the relative volatil- ities in table 7 suggests that the imidazolium-based ILs with
ity of the ester to the alcohol and eliminating the azeotrope at a hydrogen sulfate anion are the best extraction solvents for the sep-
specific IL mole fraction: 0.129 and 0.20 for methanol + methyl aration of these azeotropes.
acetate and ethanol + ethyl acetate, respectively. Ethanol + ethyl acetate azeotrope was also the most studied
Azeotrope ethanol + ethyl acetate is the most frequently studied system for liquid–liquid extraction. Figure 17 illustrates the influ-
system of this type. Figure 16 shows the influence of ILs on the rel- ence on the selectivity values of alkyl or hydroxyl substituent in
ative volatility of ethyl acetate to ethanol. The separation ability of the imidazolium cation of the ILs. These data show that [(EtOH)-
the four ILs at xIL = 0.30 is in the following order: [OMIM][BF4] > [E- MIM][BF4] has greater extraction capacity than the other ILs and
MIM][BF4] > [EMIM][OTf] > [BMIM][BF4] and is mainly due to the that a hydroxyl group on the cation and a methyl group at the
difference in polarity of the ILs and the demixing effect [102]. 2-position can obviously affect the IL ethanol interaction. The
These results show that the tetrafluoroborates-based ILs are prom- selectivity of the four ILs are ordered as follows: [(EtOH)MIM]
ising additives for the separation of ethylacetate + ethanol by [BF4] > [(EtOH)MMIM][BF4] > [EMIM][BF4] > [EMMIM][BF4]. The
extractive distillation. selectivity of [EMIM][BF4] is larger than [EMMIM][BF4], and that
However, the liquid–liquid extraction of (alcohol + ester) sys- of [(EtOH)MIM][BF4] is clearly larger than [(EtOH)MMIM][BF4].
tems using ILs as solvents has been the most commonly used sep- [(EtOH)MMIM][BF4] and [EMMIM][BF4] have a methyl instead of
aration process, which may offer a high selectivity process with hydrogen in the 2-position of the imidazolium cation greatly
 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28 17

FIGURE 16. Acetate/alcohol relative volatility versus IL [101–103] used for (etha- FIGURE 18. Experimental binodal curves of the ternary systems (2-propanol + ethyl
nol + ethyl acetate) system at azeotropic composition (xEthyl acetate  0.54) and acetate + [CnMIM][PF6]) at 298.15 K [108]: d, [BMIM][PF6]; , [HMIM][PF6]; and j,
xIL  0.30. [OMIM][PF6].

2.2.2. (Alcohols + ketones) systems

Acetone and methanol are two of the most widely used solvents
and their mixtures are frequent. The method for recovering volatile
solvents is typically by rectification in multiplate column. Com-
plete recovery is, however, impossible due to the formation of a
minimum boiling azeotrope. For example, 2-butanone has been
traditionally used as a solvent in paints and resin adhesives. Mix-
tures of different alcohols with this ketone forming azeotropes
are very common sub-products in the coating industry.
The separation of these types of azeotropes is carried out by
extractive or azeotropic distillation [109–111]) using as entrainers
either: 3-pentanone, acetophenone, 2,4-pentanedione, ethylace-
toacetate, or 2-butanone plus benzyl for the azeotrope metha-
nol + acetone; amyl acetate, methyl formate, 2,2-dimethyl butane
or 2,3-dimethyl butane for the azeotrope ethanol + 2-butanone;
and 3-methylpentane, amyl ether or acetonitrile for the azeotrope
2-propanol + 2-butanone.
FIGURE 17. Alcohol/ester selectivity values versus ILs with the [BF4] anion [105] Table 8 [112–114] summarizes data using ILs for the separation
used in the separation of the (ethanol + ethyl acetate) system at 298.15 K and of these azeotropes. Orchilles et al. [112] are the only group to have
ethanol molar fraction in organic phase  0.10.
evaluated [EMIM][OTf] as a potential entrainer for the separation
of the azeotrope methanol + acetone. They demonstrated that the
addition of IL to the azeotrope produced an important salting-out
decreasing the ability of the cation to form a hydrogen bond with effect while the azeotrope disappeared for small amounts of IL
ethanol [105]. The selectivity of [(EtOH)MIM][BF4] is clearly larger (xIL = 0.078).
than that of [EMIM][BF4], while that of [(EtOH)MMIM][BF4] is obvi- Table 9 compares [EMIM][OTf] and entrainers conventionally
ously greater than [EMMIM][BF4]. The difference between the used to separate the methanol + acetone azeotrope. The ratio and
selectivity values results from the hydroxyl group in [(EtOH)- relative volatility are not distinctly different. However, ILs are
MIM][BF4] and [(EtOH)MMIM][BF4], a fact which clearly enhances non-volatile (at the range of temperatures and pressures imple-
the interaction between the IL and ethanol (probably by hydrogen mented) demonstrating significant advantages: high-boiling and
bonding). thermal stable compounds, minimizing distillate contamination;
Pereiro et al. [108] studied [MMIM][MeSO4], [BMIM][PF6] and and the ability to regenerate non-volatile entrainers via stripping,
[HMIM][PF6] as extraction solvents in the separation of 2-propa- evaporation, drying, or crystallization.
nol + ethyl acetate azeotrope. The binodal curve for the ternary The other alcohol + ketone azeotropes were researched by li-
system (2-propanol + ethyl acetate + [OMIM][PF6]) was also deter- quid–liquid extraction. A comparison between the selectivity val-
mined. Figure 18 illustrates how the immiscibility region decreases ues for the ternary systems shows that [MMIM][MeSO4] obtains
with the increase in the length of the alkyl chain in the imidazoli- higher values than [BMIM][PF6] for the ethanol + 2-butanone and
um ring when the [CnMIM][PF6] is used as solvent in liquid–liquid the opposite is observed for 2-propanol + 2-butanone. All the selec-
extraction. The analysis of the LLE data [108] indicates that the al- tivity values are higher than unity, from which it can be inferred
kyl chain length of the imidazolium ring plays a negative role in that the extraction of the solute from the azeotropic system is in-
the extraction solvent capability of [CnMIM][PF6], due, perhaps, deed possible.
to the hydrophobic steric effect of the alkyl group reducing the po- Figure 19 illustrates binodal curves for the systems of (etha-
lar character of 2-propanol. However, the use of [MMIM][MeSO4] nol + 2-butanone + [BMIM][PF6] or [MMIM][MeSO4]), (2-propa-
as solvent leads to higher values of selectivity than those obtained nol + 2-butanone + [BMIM][PF6] or [MMIM][MeSO4]). For these
using ILs involving [PF6] as the anion, owing to the fact that the two azeotropes, [BMIM][PF6] is miscible with ketone and immisci-
[MMIM][MeSO4] IL contains just two methyl groups and the steric ble with alcohol. The opposite occurs for [MMIM][MeSO4], which is
alkyl effect is, therefore, reduced. miscible with alcohol and immiscible with ketone. This is a clear
18 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28

TABLE 8
Azeotropes in the literature regarding the separating of alcohols + ketones. ILs used as azeotrope breakers. Solute/inert selectivity from LLE data (SLLE) and acetone/methanol
relative volatility from VLE data (aVLE).

Azeotrope Ionic liquids SLLE a

aVLE b Remarks Reference
Methanol + acetone [EMIM][OTf] 1.52 VLE data at 100 kPa [112]
Ethanol + 2-butanone [BMIM][PF6] 1.64c LLE data at 298.15 K [113]
[MMIM][MeSO4] 2.81d LLE data at 298.15 K [114]
2-Propanol + 2-butanone [BMIM][PF6] 4.14c LLE data at 298.15 K [113]
[MMIM][MeSO4] 1.53d
a
Alcohol molar fraction in organic phase  0.10.
b
Azeotropic composition and IL molar fraction = 0.36.
c
2-Butanone/ethanol selectivity.
d
Ethanol/2-butanone selectivity.

TABLE 9
Ratio (volume of entrainer per volume of azeotrope) and acetone/methanol relative
volatility for entrainers effective in separating of methanol + acetone azeotrope.

Entrainers Ratio Relative volatility

a
3-Pentanone 3/4 1.41
Acetophenonea 3/5 2.27
2,4-Pentanedionea 3/5 1.89
Ethylacetoacetatea 3/5 2.17
[EMIM][OTf]b 3/5 1.52
a
From reference [109].
b
From reference [112].

FIGURE 20. Alcohol/alkane selectivity versus ILs [120–128] used in the separation
of the (alcohols + alkanes) systems at 298.15 K and alcohol molar fraction in organic
phase  0.01.

example of the tailor-made properties of the IL: a change in the an-

ion leads to a different extracted product. A comparison of the bin-
odal curves shows that the immiscibility area increases when
changing from ethanol to 2-propanol in the system (2-buta-
none + [BMIM][PF6]), and decreases slightly when the IL is changed
from [MMIM][MeSO4] to [BMIM][PF6] for the (ethanol + 2-buta-
none) system.
FIGURE 19. Experimental binodal curves at 298.15 K [113,114] of the ternary
systems: (2-propanol + 2-butanone + [BMIM][PF6]); (2-butanone + 2-propa- 2.2.3. (Alcohols + aliphatic hydrocarbons) systems
nol + [MMIM][MeSO4]); (ethanol + 2-butanone + [BMIM][PF6]); and (2- The reduction of lead in gasoline has led to an increase in the
butanone + ethanol + [MMIM][MeSO4]). number of processes under development or already in production

TABLE 10
Azeotropes in the literature regarding the separating of alcohols + alkanes. ILs used as azeotrope breakers. Alcohol/alkane selectivity from LLE data (SLLE) at 0.01 alcohol molar
fraction in organic phase.

Azeotrope Ionic liquids SLLE Remarks Reference

Methanol + heptane [OMIM][Cl] 401.0 LLE data at 298.15 K [120]
Ethanol + hexane [EMIM][EtSO4] 3631 LLE data at 298.15 K [121]
[HMM(EtOH)N][BF4] 1584 LLE data at 298.15 K [122]
[BMIM][MeSO4] 482.1 LLE data at 298.15 K [123]
[MMIM][MeSO4] 9427 LLE data at 298.15 K [124]
[HMIM][PF6] 330.7 LLE data at 298.15 K [125]
[OMIM][PF6] 115.2
Ethanol + heptane [EMIM][EtSO4] 2343 LLE data at 298.15 K [121]
[MMIM] [MeSO4] 5592 LLE data at 298.15 K [126]
[OMIM][PF6] 229.8 LLE data at 298.15 K [125]
[BMIM][MeSO4] 22,326 LLE data at 298.15 K [127]
[HMIM][PF6] 2861 LLE data at 298.15 K [128]
[OMIM][Cl] 457.0 LLE data at 298.15 K [120]
 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28 19

FIGURE 21. Mass composition (%) of the feed, raffinate and extract streams of the packed column under steady state conditions for ([MMIM][MeSO4] [124,126] and
[BMIM][MeSO4] [123,127].

TABLE 11 compounds from the extract to regenerate the ILs used in this
Extraction efficiencies (from equation (6)) for the separation of the azeotrope extraction process. Visible inspection of the NMR spectra and phys-
ethanol + alkanes in the packed column [123,124,126,127,129]. ical properties of both the original IL and the regenerated product
Azeotrope [MMIM][MeSO4] [BMIM][MeSO4]
showed no differences between the two samples. This means that
no degradation of the IL after repeated recycling was confirmed.
Ethanol + hexane 1.81 1.14
Ethanol + heptane 1.26 1.13
The performance of the experimental column exceeds the theo-
retical (from (liquid + liquid) equilibria data) and simulation-based
expectations, certainly due to the superior mixing of the feed and
[115] in which alkanols and alkanes are brought together to pro- the solvent. Figure 21 compares the compositions of the initial feed
duce oxygenated additives for gasolines. The azeotropes of either and the two outlet streams of the countercurrent packed column
hexane or heptane with methanol or ethanol are published due for the two ILs ([MMIM][MeSO4] and [BMIM][MeSO4]). The results
to difficulties in separating these compounds. clearly indicate that [MMIM][MeSO4] has less alkane in the extract
The liquid–liquid separation leads to an environmentally stream. In order to emphasize the two ILs for extraction processes
friendly extraction process of these azeotropes as an alternative in the packed column, the extraction efficiency, EEP, was calculated
to the more common processes such as azeotropic distillation (from equation (6)). This parameter indicates the ability of the ILs
[116,117], pervaporation [118], and reverse osmosis [119]. Table to remove solute (ethanol) from the azeotropes (ethanol + alkane)
10 [120–128] evaluates IL azeotrope breakers for alka- in the extraction column. Table 11 lists the extraction efficiencies
nols + alkanes. Figure 20 compares alcohol/alkane selectivity val- of the ILs in the separation of the ethanol + alkanes (hexane and
ues for the studied ternary systems for the ILs listed in table 10. heptane) azeotropes. The values obtained for [MMIM][MeSO4]
It can be concluded that a short alkyl side chain on the imidazoli- are higher than those for [BMIM][MeSO4]. Moreover, [MMIM][Me-
um cation increases selectivity, favoring ethanol/alkane separation. SO4] has both lower viscosity and synthesized cost.
Based on the analysis of the ILs extraction capacities for etha-
nol + alkanes (hexane or heptane) systems, ILs with the methyl sul- 2.2.4. (Alcohols + halogenated) systems
fate anion ([MMIM][MeSO4] and [BMIM][MeSO4]) were included in Nowadays, great importance is given to the separation of organ-
a lab-scale extraction process incorporating a solvent recycling ic waste composed mainly of chloroform and methanol, a widely
stage [123,124,126,127,129]. These selectivities are the highest re- used effective solvent mixture for the extraction of bioactive sub-
ported in the literature for azeotropic separation in general using stances from biological sources in the pharmaceutical and biotech-
liquid–liquid extraction. The (liquid + liquid) equilibria data al- nological industry [130–132]. This binary system exhibits a
lowed researchers to identify theoretically appropriate operating minimum boiling temperature azeotrope xChloroform = 0.648 at
conditions for countercurrent continuous extraction process at 328.15 K and atmospheric pressure (table 1). Then, non-conven-
room temperature including a solvent recycling stage. Simulation tional distillation processes as extractive or azeotropic distillation
techniques then optimized operational conditions. Experiments are required to separate both components. In addition to its indus-
with a laboratory-scale packed column under steady-state condi- trial importance, theoretical issues regarding the interesting
tions achieved a raffinate purity of over 98 wt%. behavior verified in systems containing chloroform and lower ali-
The lab-scale experiments also confirmed the possibility of phatic alcohols were first addressed by Scatchard et al. [133]
ready online recovery of selected ILs, by evaporating the organic through VLE measurements.

TABLE 12
Azeotropes in the literature regarding the separating of alcohols + halogenated. ILs used as azeotrope breakers. Alcohol/chloroform selectivity from LLE data (SLLE) and chloroform/
alcohol relative volatility from VLE data (aVLE) at azeotropic composition.

Azeotrope Ionic liquids SLLE a

aVLE Remarks Reference
Methanol + chloroform [EMIM][OTf] 5.42 1.62b VLE data at 100 kPa + LLE data at 293.15 K [134]
Ethanol + chloroform [EMIM][OTf] 3.62 1.82c VLE data at 100 kPa + LLE data at 293.15 K [135]
a
Alcohol molar fraction in organic phase  0.01.
b
Azeotropic composition and IL molar fraction = 0.35.
c
Azeotropic composition and IL molar fraction = 0.32.
20 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28

FIGURE 23. Influence of the [EMIM][OTf] crossover effect on the selectivity for both
salting-out (xChloroform = 0.9407) and salting-in (xChloroform = 0.1081) regions
FIGURE 22. Influence of [EMIM][OTf] on the VLE of (chloroform + methanol) [134,135]. The straight horizontal line represents the selectivity value, equal to
system at 100 KPa for different IL concentrations [134]. unity, when the relative volatility is equal with and without the entrainer.

The ILs reported in table 12 [134,135] were used as azeotrope

breakers for the (alcohols + chloroform) systems using extractive 2.3. (Aromatic + aliphatic hydrocarbons) systems
distillation and liquid–liquid extraction.
Orchiles and co-workers [134,135] presented the effect of [EMI- Several processes [142] for the separation of aromatic and ali-
M][OTf] on the VLE for the systems (chloroform + methanol) and phatic hydrocarbon mixtures in reformate gasoline distilled from
(chloroform + ethanol). Both systems exhibit, at atmospheric pres- crude have been developed: liquid extraction for aromatic contents
sure, a minimum boiling temperature azeotrope at xChloro- from 20 to 65 wt%, extractive distillation for 65 to 90 wt%, and aze-
form = 0.648 for the methanol system and xChloroform = 0.841 for otropic distillation for aromatic content higher than 90 wt%. These
the ethanol system (table 1). For these systems, inorganic salts products are the most widespread and important raw materials in
such as lithium chloride [136,137], calcium chloride [137,138], cal- plastic and synthetic fiber manufacture [143]. Aromatics are ob-
cium nitrate [139], potassium acetate [137], and sodium iodide tained almost exclusively from fuels, coal, and oil; however, the
[139,140] have been used, rather unsuccessfully, as entrainers to most economically feasible process is to remove them from fuels.
break the azeotrope mostly due to the low solubility of these salts Typical solvents used in the separation of (aromatic + aliphatic)
at large chloroform concentrations. These authors demonstrated mixtures, through liquid–liquid extraction, are sulfolane [144–
the capacity of [EMIM][OTf] to break both azeotropes and its apt- 146], N-methylpyrrolidone [144,147], N-formylmorpholine
ness as entrainer in extractive distillation is clearly elucidated in [146,148], ethylene glycols [149,150], and propylene carbonate
figure 22. [151]. When these conventional solvents are used, additional dis-
Analysing the displacement of azeotropic point depicted in fig- tillation steps are required to separate the solvent from both the
ure 22 as a function of [EMIM][OTf] concentration, it is readily ob- extract and raffinate phases and to purify the solvent, increasing
served that, for chloroform + methanol at a IL composition of 0.23, separation costs. The costs of regeneration of sulfolane are high,
the azeotrope has not yet disappeared, but for the chloro- because, in the current process, sulfolane, which has a boiling tem-
form + ethanol system (supporting information figure S9) at a IL perature of 287.3 °C, is taken as overhead from the regenerator and
composition of 0.21 the azeotrope has done so. This is explained returned to the bottom of the aromatics stripper as a vapor [152].
by the fact that the chloroform relative volatility is higher in the Most ethylene cracker feeds contain 10 to 25 wt% aromatic
(chloroform + ethanol) system than in the (chloroform + methanol) components, depending on the source of the feed (naphtha or
system. Furthermore, since methanol and ethanol possess similar gas condensate). The aromatic compounds are not converted into
dipole moments, it is obvious that the (chloroform + methanol) olefins, and even small amounts are formed during the process in
system will need a greater electrolyte concentration to the the cracker furnaces. Therefore, these compounds partly occupy
azeotrope. the furnace capacity, putting an extra load on the separation sec-
Both methanol and ethanol are more polar than chloroform. tion of the stream containing C5–C10 aliphatic compounds. If a sig-
Hence, the latter is supposed to be salted-out from the mixture nificant part of the aromatic compounds present in the feed to the
over the whole range of liquid concentration. However, figure 22 crackers could be separated upstream, it would offer several
demonstrates that the salting-out effect is only verified for chloro- advantages: higher capacity, higher thermal efficiency, and less
form compositions higher than 0.4 for the system with methanol, fouling [153]. The improved margin for the removal of 10 wt%
and higher than 0.5 for the system with ethanol (supporting infor- aromatic hydrocarbons from the feed to the naphtha cracker would
mation figure S9). On the other hand, a salting-in effect occurs at represent a savings of around € 20/t of feed or € 48 million per year
lower compositions. [EMIM][OTf] produces a cross-over effect for a cracker with a feed capacity of 300 t/h [153]. Although the
[141], a fact clearly demonstrated in figure 23 by the selectivity separation of aromatic and aliphatic hydrocarbons after the
values for both salting-in and salting-out regions in (chloro- furnace section is current practice, no suitable technology yet
form + methanol + [EMIM][OTf]) and (chloroform + ethanol + exists for separating aromatic compounds from the feed to
[EMIM][OTf]) systems. cracker plants [153]. Different ILs for the separation of
Orchilles and co-workers [134,135] also researched the separa- (aromatic + aliphatic hydrocarbons) azeotropic systems were
tion of the alcohol + chloroform azeotropes by liquid–liquid extrac- therefore tested to wAromatic = 0.10 in organic phase.
tion using [EMIM][OTf] as solvent. These systems present small Table 13 [30,65,153–179] lists the aromatic/aliphatic hydrocar-
immiscibility regions, increasing with temperature at the highest bons azeotropes using ILs as azeotrope breakers. These research
chloroform compositions. The highest alkanol/chloroform selectiv- data study the azeotropic separation by extractive distillation (3
ities for [EMIM][OTf] were obtained at low concentration of alcohol papers), liquid–liquid extraction (26 papers) and supported liquid
in organic phase (supporting information figure S10). membrane (4 papers).
 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28 21

TABLE 13
Azeotropes in the literature the separating of aromatic + aliphatics hydrocarbons. ILs used as azeotrope breakers. Aromatic/aliphatic selectivity from LLE data (SLLE). Aliphatic/
aromatic relative volatility from VLE data (aVLE). Aromatic/aliphatic selectivity from membranes data (SMEM) and aromatic/aliphatic separation factor from membranes data
(aMEM).

Azeotrope Ionic liquids SLLE a

aVLE b SMEM aMEM Remarks Reference
Benzene + hexane [HMIM][NTf2] 9.49 VLE data at 322.8 K: headspace gas chromatography [30]
[OMIM][NTf2] 7.53
[EMIM][N(CN)2] 30.10 LLE data at 303.15 K [153]
[EMIM][SCN] 114.3
[BMIM][N(CN)2] 30.50
[BMIM][C(CN)3] 32.30
[BMIM][SCN] 55.40
[HMIM][SCN] 20.90
[3MBpy][BF4] 36.20
[3MBpy][N(CN)2] 35.30
[3MBpy][C(CN)3] 34.80
[3MBpy][B(CN)4] 27.00
[4MBpy][N(CN)2] 34.30
[4MBpy][SCN] 45.20
[BMpyr][NTf2] 20.70
[BMpyr]][N(CN)2] 15.40
[BMpyr][SCN] 17.30
[MOOON][NTf2] 1.80
[EEES][NTf2] 32.20
[3MBpy][N(CN)2] 48.88 LLE data at 303.15 K [154]
[BMIM][MeSO4] 57.99 LLE data at 298.15 K [155]
[EPy][EtSO4] 62.30 LLE data at 298.15 K [156]
[EMIM][EtSO4] 48.55 LLE data at 298.15 K [157]
[EMPy][EtSO4] 26.12 LLE data at 298.15 K [158]
[EMIM][EtSO4] 87.68 LLE data at 298.15 K [159]
[Epy][NTf2] 27.70 LLE data at 313.15 K [160]
[MMM(EtOH)N][NTf2] 41.90
[P6 6 6 14][NTf2] 1.07 LLE data at 298.15 K [161]
[EMIM][NTf2] 27.90 LLE data at 298.15 K [162]
[BMIM][NTf2] 15.89 LLE data at 298.15 K [163]
[OMIM][NTf2] 7.35
[DMIM][NTf2] 4.90
[DoMIM][NTf2] 3.54
[EMM(EtOH)N][NTf2] 20.51 LLE data at 298.15 K [164]
[4MBpy][BF4] 55.01 LLE data at 313.15 K [165]
Benzene + heptane [EMIM][EtSO4] 71.66 LLE data at 298.15 K [157]
[BMIM][BF4] 110.0 LLE data at 298.15 K [166]
[BMIM][SCN] 95.63
[MMIM][MePO3] 843.2
[BMpyr][NTf2] 17.91 LLE data at 298.15 K [167]
[EPy][EtSO4] 67.27 LLE data at 298.15 K [156]
[EMPy][EtSO4] 71.60 LLE data at 298.15 K [158]
[BMIM][Cl] – 2.0 AlCl3 80.00 LLE data at 293.15 K [168]
[MMMNH][Cl] – 2.0 AlCl3 38.00
[EEENH][Cl] – 2.0 AlCl3 40.20
[EEENH][Cl] – 1.5 AlCl3 32.00
[EEENH][Cl] – 1.2 AlCl3 28.00
[EMIM][OcSO4] 2.71 LLE data at 298.15 K [169]
[OMIM][Me(EtO)2SO4] 0.63
[HMIM][BF4] 8.35 LLE data at 298.15 K [170]
[HMIM][PF6] 11.20
[OMIM][Cl] 6.06 LLE data at 298.15 K [171]
[BMIM][PF6] 67.00 Supported liquid membranes: permeation [172]
[HMIM][PF6] 16.00
[OMIM][PF6] 7.50
[EEM(MeOEt)N][NTf2] 7.20
Benzene + cyclohexane [OMIM][NTf2] 3.93 VLE data at 322.8 K: headspace gas chromatography [30]
[HMIM][NTf2] 5.04 VLE data at 332 K [65]
[BMpyr][NTf2] 7.75
[EMpy][EtSO4] 26.97 LLE data at 298.15 K [173]
[4MBpy][BF4] 28.25 LLE data at 303.15 K [174]
[MIM][(Bu)2PO4] 1.90 LLE data at 298.15 K [175]
[MMIM][(Me)2PO4] 3.60 LLE data at 298.15 K [176]
[EMIM][(Et)2PO4] 4.30
[BMIM][PF6] 11.60 Supported Liquid Membranes: Pervaporation [177]
[HMIM][PF6] 10.30
[OMIM][PF6] 2.50
[EEM(MeOEt)N][NTf2] 10.90
[EEM(MeOEt)N][BF4] 47.10
[BMIM][PF6] 19.70 Supported liquid membranes: vapor permeation [178]
[HMIM][PF6] 10.80

(continued on next page)

22 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28

TABLE 13 (continued)

Azeotrope Ionic liquids SLLE a

aVLE b SMEM aMEM Remarks Reference
[OMIM][PF6] 5.90
[EEM(MeOEt)N][NTf2] 6.92
[EEM(MeOEt)N][BF4] 185.0
Pyridine + heptane [BMIM][PF6] 9.00 Supported liquid membranes: permeation [179]
[HMIM][PF6] 7.60
[OMIM][PF6] 4.90
[EEM(MeOEt)N][NTf2] 3.80
[EEM(MeOEt)N][BF4] 8.60
a
Aromatic mass fraction in organic phase  0.10.
b
Azeotropic composition and IL molar fraction  0.81.

FIGURE 24. Benzene/hexane selectivity versus benzene distribution coefficient for the ILs used in the separation of (benzene + hexane) [153–155] at 298.15 K and benzene
mass fraction in organic phase  0.10. [EMIM][EtSO4] (a) from reference [159] and [EMIM][EtSO4] (b) from reference [157]; [3MBpy][EtSO4] (a) from reference [153] and
[3MBpy][EtSO4] (b) from reference [154].

2.3.1. (Benzene + hexane) system 2.3.1.1. Effect of the alkyl chain length. Figure 25 illustrates the evo-
While most research has focussed on liquid–liquid extraction, lution of the benzene/hexane selectivity and benzene distribution
Mokhtarani et al. [30] studied the separation the azeotrope (ben- coefficient as a function of the cation alkyl chain length in the IL
zene + hexane) by extractive distillation. These results showed that [CnMIM][NTf2]. The use of the IL with the shortest alkyl chain,
the addition of [HMIM][NTf2] and [OMIM][NTf2] as entrainers im- n = 2 ([EMIM][NTf2]), leads to the highest selectivity values. For
proves separation. the distribution coefficient, however, [EMIM][NTf2] has the lowest
The system (benzene + hexane) is representative of aromatic/ values, which rise with moderate increases in the length of the al-
aliphatic hydrocarbon separation by liquid–liquid extraction. IL kyl chain of the IL. The increment in the distribution coefficient is
criteria were: a comparable or higher benzene distribution coeffi- smaller than the decline in selectivity values when the IL chain of
cient and benzene/hexane selectivity than sulfolane (0.35 mol/ the IL lengthens.
mol and 29.1 at 303.15 K and wAromatic = 0.10 in organic phase, Benzene is known to interact with ILs such as [EMIM][NTf2] to
respectively) [153]. Temperature has been shown to have little form an inclusion compound, [EMIM][NTf2]–C6H6 [180]. The resul-
influence on the phase equilibria of this kind of system tant stability is derived from the anions clustering around the ben-
[160,162]. Figure 24 plots benzene/hexane selectivity as a function zene and derives from interactions between the benzene hydrogen
of the distribution coefficient (mol/mol) of benzene compared with atoms and the oxygen atoms of the sulfonyl groups in the anions
the sulfolane values. The most suitable ILs for the separation of [180]. The compound decreases benzene–hexane (azeotrope)
benzene and hexane are: [3MBpy][C(CN)3] with a distribution interactions, facilitating separation. Cation benzene interactions
coefficient of 1.84 (mol/mol) and a selectivity of 34.8; are mainly due to p–p stacking which promotes quadrupolar ben-
[BMIM][C(CN)3] with a distribution coefficient of 1.54 (mol/mol) zene molecule associates with the anion and the imidazolium part
and a selectivity of 32.3; [3MBpy][N(CN)2] with a distribution coef- of the cation.
ficient of 1.44 (mol/mol) and a selectivity of 35.3 [153] or a distri- The distribution coefficients (figure 25) at low benzene concen-
bution coefficient of 1.19 (mol/mol) and a selectivity of 48.9 [154]; trations in organic phase (wAromatic = 0.10) presented values in the
and [3MBpy][B(CN)4] with a distribution coefficient of 1.92 (mol/ following order: C12 < C10  C8 > C4 > C2. A possible explanation is
mol) and a selectivity of 27.0. All values are higher than those ob- that hexane is moderately soluble in [DoMIM][NTf2] (n = 12)
tained with sulfolane. [163]. The increasing hexane concentration in the IL as the chain
 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28 23

FIGURE 25. Effect of cation alkyl chain length on benzene/hexane selectivity and
benzene distribution coefficient in ILs [CnMIM][NTf2] [162,163] at 298.15 K and
FIGURE 26. Effect of the anion nature of ILs with the [EMIM]+ or [BMIM]+ cation on
benzene mass fraction in organic phase  0.10.
the benzene/hexane selectivity for the benzene + hexane azeotrope [153,155,157,
159,162,163] at benzene mass fraction in organic phase  0.10 (for [EtSO4] anion:
(a) from reference [159] and (b) from reference [157]).
length increases lowers the distribution ratio [163]. The best distri-
bution ratios occur for ILs that are immiscible with benzene. Short
chain ILs, such as [EMIM][NTf2], can dissolve benzene quite well,
but, for hexane the anion–cation interactions would be disrupted
if they were to mix. For long chain [CnMIM][NTf2] ILs, hexane can
be accommodated in the hydrophobic region of the alkyl side
chains and hence its solubility in the IL phase improves [163]. Ben-
zene can also be accommodated in these nonpolar regions, and
consequently its solubility increases to such an extent that it is
completely miscible with [DoMIM][NTf2] [163].

2.3.1.2. Effect of the nature of the anion. Figure 26 depicts the ben-
zene/hexane selectivity regarding the separation of the azeotrope
benzene + hexane using various ILs with differing anions as sol-
vents. These ILs are based on the [EMIM]+ and [BMIM]+ cations
which were combined with six different anions.
The effect of the anion on selectivity follows the trend for the
[EMIM]+ cation: [SCN] > [EtSO4] > [N(CN)2]  [NTf2]; and for
FIGURE 27. Effect of the cation nature of ILs with the [N(CN)2] or [NTf2] anion on
the [BMIM]+ cation: [MeSO4]  [SCN] > [C(CN)3]  [N(CN)2] > the benzene/hexane selectivity for the (benzene + hexane) azeotrope [153,154,160–
[NTf2]. Generally, anions with delocalized charge and with small 164] at benzene mass fraction in organic phase  0.10 (for [3MBpy]+ cation: (a)
molar volume seem to be efficient IL solvents. Especially [MeSO4], from reference [153] and (b) from reference [154]).
[SCN], and [EtSO4] are the best candidates for forming an effec-
tive solvent for this liquid–liquid extraction. Anions with delocal-
ized charges interact less with cations, so benzene interactions Moreover, in the case of the [MMM(EtOH)N]+ cation system, its hy-
with cations are favored. Furthermore, with low molecular droxyl group is likely to favorably interact with the aromatic ring
weights, the number of IL molecules in the vicinity of the benzene of benzene, while a cation-p interaction may also exist [160].
molecule is high because of its low molar volume. The result is in-
creased benzene selectivity with these ILs. 2.3.2. (Benzene + heptane) system
Subsequent to the (benzene + hexane) system, benzene + hep-
2.3.1.3. Effect of the nature of the cation. Figure 27 depicts benzene/ tane can be used to model aromatic/aliphatic hydrocarbons sepa-
hexane selectivity as a function of the nature of the cation on the rations. The benzene distribution coefficient for sulfolane is
separation of the azeotrope benzene + hexane, using as solvent 0.46 mol/mol and the benzene/heptane selectivity is 22.2 at
various ILs with the dicyanamide or bis(trifluoromethylsulfo- 303.15 K and wAromatic = 0.10 in organic phase [181]. The IL selec-
nyl)imide anions. These selectivities at low benzene concentra- tion criteria were those for where the benzene distribution coeffi-
tions in organic phase (wAromatic = 0.10) present values in the cient is greater then 0.46 mol/mol and benzene/hexane selectivity
following order: [3MBpy]+ > [4MBpy]+ > [EMIM]+  [BMIM]+ > greater than 22.2, at T  298.15 K and at wBenzene = 0.10 in organic
[MBpyr]+ for the [N(CN)2] anion; and [MMM(EtOH)N]+ > phase (comparison between ILs and sulfolane in supporting infor-
[EEES]+ > [Epy>]+  [EMIM]+ > [MBpyr]+  [EMM(EtOH)N]+ > [BMI- mation figure S11). Room temperature was considered as the refer-
M]+ > [MOOON]+  [P6 6 6 14]+ for the [NTf2] anion. Again, the re- ence due to the research tendency to determine the
ported benzene/hexane selectivities in the abovementioned ILs (liquid + liquid) equilibria of ternary systems at this temperature,
decrease with increasing alkyl chain length of the [CnMIM]+ cation. given that all compounds and their mixtures are liquid at such
This behavior is observed in ammonium-based, imidazolium- temperature.
based and pyridinium-based ILs. Table 13 lists the ILs suitable for benzene/heptane separation.
For the same [N(CN)2] anion and similar alkyl chain length, The chloroaluminate IL [BMIM][Cl]–2.0 AlCl3 shows the highest
pyridinium ILs have higher selectivities than imidazolium ILs. This benzene distribution coefficient and the highest benzene/heptane
can be probably attributed to the higher aromatic character of selectivity (supporting information figure S11), but this IL is not
pyridinium. This behavior is also observed for the [NTf2] anion. suitable for industrial processes due to its reaction with water
24 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28

collects the ionic liquids used as azeotrope breakers for this system
together with the values of the parameters for each process with
the aim of evaluating separation efficiency. The best extractive dis-
tillation entrainer IL is [BMpyr][NTf2]. In the case of liquid–liquid
extraction, [EMpy][EtSO4] and [4MBpy][BF4] present the highest
values of benzene/cyclohexane selectivity.
Figure 28 demonstrates the results for the separation of the
benzene + cyclohexane or heptane and pyridine + heptane using
supported liquid membranes. These data show that the use of ILs
as a membrane solvent enables this technique to be used for the
separation of aromatic and aliphatic hydrocarbons. Aromatic
hydrocarbons were successfully transported through the mem-
brane. The hydrophilic liquid membrane that used [EEM(MeO-
Et)N][BF4] led to the highest separation factor for the azeotrope
benzene + cyclohexane, 185 for 53 wt% benzene obtained by vapor
permeation, which is superior to the reported values for pervapo-
ration (selectivity of 47.1 for the mixed solution). Other good re-
FIGURE 28. Selectivity or separation factor versus ILs used in the separation of the sults were obtained for the azeotrope benzene + heptane. The
(aromatics + aliphatic) systems by supported liquid membranes [172,177–179]. maximum selectivity to heptane resulted from the use of benzene
in the aromatic permeation and [BMIM][PF6] in the liquid mem-
[153]. The ILs [MMMNH] [Cl]–2.0 AlCl3 and [EEENH] [Cl]–2.0 AlCl3 brane phase (selectivity of 67.0 for the mixed solution).
(or 1.5 AlCl3 or 1.2 AlCl3) are not suitable for the same reason
(protic ILs). In the case of [BMIM][BF4] IL, with its high selectivity
value, the anion [BF4] has been commonly investigated and con- 2.4. Other azeotropes
sidered historically to be one of the most important anion families.
However, this anion can undergo hydrolysis producing hydroflu- 2.4.1. (Methanol + dimethyl carbonate) system
oric acid when in contact with water and at high temperatures Dimethyl carbonate has been used as an effective carbonylation
[182,183]. Consequently, its industrial application has been re- agent to substitute phosgene in manufacturing polycarbonate plas-
stricted to those processes under water-free conditions and mod- tics and as a useful methylation agent to replace the dimethyl sul-
erate temperatures. fate [184]. Moreover, dimethyl carbonate has been successfully
The most suitable ILs for the separation of benzene and heptane manufactured using carbon dioxide, the greenhouse gas, and
are: [BMIM][SCN] with a distribution coefficient of 1.03 (mol/mol) methanol as raw materials according to a zero-discharge process
and a selectivity of 95.6; [EMpy][EtSO4] with a distribution coeffi- co-producing propylene glycol [185]. As a result, with respect to
cient of 0.80 (mol/mol) and a selectivity of 71.6; [EMIM][EtSO4] the carbonylation usage of dimethyl carbonate, phosgene is simply
with a distribution coefficient of 0.56 (mol/mol) and a selectivity replaced with carbon dioxide, which may be one of the most signif-
of 71.7; and [BMpyr][NTf2] with a distribution coefficient of 1.25 icant improvements in the chemical industry. Dimethyl carbonate
(mol/mol) and a selectivity of 17.9. All these values are higher than is prepared from the catalytic oxidative carbonylation of methanol
those obtained with sulfolane for the separation of the azeotrope and forms an azeotrope with methanol. The separation of this
benzene + heptane. azeotrope has been investigated by extractive distillation [186]
and/or by the use of membranes [187,188].
Kim et al. [64] reports the influence of [EMIM][EtSO4] and
2.3.3. Other (aromatic + aliphatic hydrocarbons) systems [BMIM][BF4] on the separation of the azeotrope dimethyl carbon-
The azeotrope (benzene + cyclohexane) system was studied ate + methanol (table 14, [64,189–193]). Both ILs were effective
using three separation processes: liquid–liquid extraction, in breaking the azeotropic point, by increasing the mole fraction
extractive distillation and supported liquid membranes. Table 13 of methanol in the vapor phase. The azeotrope disappeared at

TABLE 14
Azeotropes in the literature regarding the separating of other azeotropes. ILs used as azeotrope breakers. Selectivity from LLE data (SLLE) and relative volatility from VLE data
(aVLE).

Azeotrope Ionic liquids SLLE a

aVLE Remarks Reference
Methanol + dimethyl carbonate [EMIM][EtSO4] 2.05b VLE data at 333.15 K: headspace gas chromatography [64]
[BMIM][BF4] 1.71b
Ethanol + ethyl tert-butyl ether [EMIM][EtSO4] 251.7c 3.42d VLE data at 101.32 kPa + LLE data at 298.15 K [189]
[EMIM][OTf] 47.30c LLE data at 298.15 K [190]
[EMIM][MeSO3] 1095c
[BMIM][ OTf] 20.28c LLE data at 298.15 K [191]
Hexane + ethyl acetate [HMIM][PF6] 186.2e LLE data at 298.15 K [192]
[OMIM][PF6] 83.29e
Cyclohexane + 2-butanone [HMIM][PF6] 114.9f LLE data at 298.15 K [193]
[OMIM][PF6] 55.30f
a
Solute molar fraction in organic phase  0.01.
b
Methanol/dimethyl carbonate relative volatility at azeotropic composition and IL molar fraction = 0.20.
c
Solute = ethanol.
d
Ethyl tert-butyl ether/ethanol relative volatility at azeotropic composition and IL molar fraction = 0.31.
e
Solute = ethyl acetate.
f
Solute = 2-butanone.
 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28 25

FIGURE 29. Selectivity (SVLE) for the ILs [27,34,37,62,63,69,71] used as entrainers in
the aqueous azeotropes separated by extractive distillation at THF or alcohol molar
fraction  0.95 and xIL  0.10.
FIGURE 30. [EMIM][EtSO4] selectivity (SLLE) versus azeotropes separated by liquid-
liquid extraction using this IL as solvent [121,157,189] at 298.15 K and solute molar
fraction in organic phase  0.05.
xIL = 0.1. There is no significant difference between the two ILs, but
[EMIM][EtSO4] showed a slightly better salting-out effect.
indicating that the former is a better choice as solvent for these
2.4.2. (Ethanol + ethyl tert-butyl ether) system azeotropic separations. Also, the increase of alkyl chain length of
The production of tertiary ethers at an industrial scale is mainly the imidazolium decreases the IL’s capacity to separate the two
carried out by the reaction of an olefine and an alcohol. Specifically, azeotropes with hexafluorophosphate.
ethyl tert-butyl ether (ETBE) is obtained by the reaction of isobu-
tene and ethanol. Due to the required excess of ethanol in the reac- 2.5. Final remarks
tor, the output stream of the unit is a mixture of ETBE and
unreacted ethanol. Although both compounds have an anti-knock Mapping azeotropes as a function of the ILs used as separation
effect, blending properties of ETBE are largely better than those agents depicted in figure 1, identified the three most utilized as
of ethanol [190]. So far, water has been the solvent most commonly azeotrope breakers: [EMIM][OTf], [EMIM][EtSO4], [BMIM][PF6],
used in purification stages of ETBE by liquid extraction [190], due [HMIM][PF6], and [OMIM][PF6] with nine azeotropes; [BMIM][BF4]
to its low cost and physical properties. with seven azeotropes; and [BMIM][Cl] with six azeotropes. ILs
Arce et al. [189–191] demonstrate the ability of the ILs based on based on [PF6] and [BF4] anions have received the most research
imidazolium cations to act as extraction solvents for liquid–liquid attention. Nevertheless, in the last decade, the use of these ILs has
extraction and as azeotrope breakers for extractive distillation (ta- slowed down since they hydrolise in water and at high tempera-
ble 14) in order to separate azeotrope ETBE + ethanol, thus purify- ture [182,183] leading to the formation of HF; their use has conse-
ing the tertiary ether, which is the most common additive for quently been restricted to those systems under water-free
improving gasoline octane indexes. A comparison between etha- conditions and moderate temperatures.
nol/ETBE selectivity values for all the ILs used as solvents in li- Most research dealing with IL entrainer efficiency in extractive
quid–liquid extraction (supporting information figure S12) shows distillation has focused on aqueous azeotropic systems. Five such
that the use of [EMIM][MeSO3] as the leads to the highest selectiv- systems illustrate the effect of ILs on the VLE behavior so that
ity values. Moreover, the data depicted in figure S12 indicates that the feasibility of extractive distillation technique can be evaluated:
the fluorination of the anion plays a negative role in the capability THF + water; ethanol + water; 1-propanol + water; 2-propa-
of the IL to purify the ether by liquid extraction. nol + water; and 2-methyl-2-propanol + water. Figure 29 depicts
the influence of the systematic cation and anion variation of imi-
2.4.3. (Hexane + ethyl acetate) and (cyclohexane + 2-butanone) dazolium-based ILs with respect to their entrainer efficiency in
systems extractive distillation for these systems: THF and 2-methyl-2-pro-
Pereiro et al. [192,193] chose the ILs [HMIM][PF6] and panol are the least polar organic solvents and, as can be verified by
[OMIM][PF6] for the separation of the azeotropic systems the selectivities values, they are largely salted-out from the aque-
(hexane + ethyl acetate) and (cyclohexane + 2-butanone) (table ous phase. [EMIM][OTf] and [BMIM][Cl] have received the most
14) by liquid–liquid extraction. These azeotropes are present in vigorous testing as entrainers in extractive distillation. However,
the process of purification of grafted polyolefins [194]. The hex- in this separation process, the best results are obtained with ILs
ane + ethyl acetate azeotrope was separated by azeotropic batch based in [Cl] and [OAc] anions. Moreover, the imidazolium salts
distillation [195] with heterogeneous entrainers such as methanol, containing these anions are superior in terms of selective interac-
acetonitrile, water, and nitromethane. The other system, cyclohex- tions and the increasing relative volatility of the organic solvents.
ane + 2-butanone azeotrope, was separated with a dephlegmation This behavior is in accordance with the fact that the [OAc] and
membrane [196]. [Cl] anions show the highest nucleophilicity, e.g., ability to hydro-
From LLE data for the ternary systems (cyclohexane + 2-buta- gen bond, of the anions studied. These anions interact with water
none + [HMIM][PF6] or [OMIM][PF6]), the experimental tie-lines forming hydrogen bonds, thus decreasing their water activity
(supporting information, figure S13) show a positive slope where and, consequently, increasing the relative volatility of the organic
the solute goes preferentially to the solvent-rich phase. Another solvents.
significant aspect is the fact that the IL does not enter in the organ- Ionic liquids based on the alkyl sulfate anion show the greatest
ic-rich phase, e.g., the presence of [HMIM][PF6] and [OMIM][PF6] potential as solvents in separation by liquid–liquid extraction of
was not detected in the upper phase. The selectivity values for the azeotropes reviewed. [EMIM][EtSO4] is the most frequently
the studied ternary systems (supporting information, figure S14) studied (figure 1) in this extraction process. Figure 30 compares
are higher for the [HMIM][PF6] than for the [OMIM][PF6], [EMIM][EtSO4] selectivities for different azeotropes separated by
26 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28

liquid–liquid extraction. The best results were obtained for the Acknowledgements
separation of ethanol with heptane or hexane. From the data pre-
sented in figure 20 and tables 10 and 11, it can be concluded that A.B.P. acknowledges Marie Curie Actions – Intra-European Fel-
[MMIM][MeSO4] shows better results than [EMIM][EtSO4] and lowships (IEF) for a Contract under FP7-PEOPLE-2009-IEF –
[BMIM][MeSO4]. 252355 – HALOGENILS. J.M.M. Araújo gratefully acknowledges
[MMIM][MeSO4] and [BMIM][MeSO4] were used as azeotrope the financial support from FCT/MCTES (Portugal) through Post-
breakers in a laboratory-scale liquid–liquid extraction process Doc Grant SFRH/BPD/65981/2009. J.M.S.S.E. and I.M.M. acknowl-
and the results also confirmed that [MMIM][MeSO4] has the high- edge Fundação para a Ciência e Tecnologia for a contract under Prog-
est extraction efficiency [123,124,126,127,129]. These papers con- rama Ciência 2007. We also wish to thank the Fundação para a
firm the assumption that the IL regeneration and recycling are Ciência e Tecnologia through Projects: PTDC/QUI/72903/2006,
indeed simple for this process, which is a great advantage relative PTDC/EQU-EQU/102949/2008, and PTDC/EQU-FTT/116015/2009.
to traditional extraction solvents.

3. Conclusions and future work Appendix A. Supplementary data

This work reviews the various processes used in the separation Supplementary data associated with this article can be found, in
of azeotropes. So far, three such processes have implemented ILs as the online version, at doi:10.1016/j.jct.2011.05.026.
azeotrope breakers: liquid–liquid extraction, extractive distillation
and supported liquid membranes. The first two have received the References
most research attention.
Ionic liquids represent a promising class of neoteric, ‘‘green’’ [1] W.M. Haynes, David R. Lide, CRC Handbook of Chemistry and Physics, 91st ed.,
engineering liquids for separating processes due to their nature Internet Version, 2011.
[2] R.D. Rogers, K.R. Seddon, Ionic Liquids – Industrial Applications to Green
as highly selective, nonvolatile azeotrope breakers or azeotrope Chemistry, ACS Symposium Series 818, Washington, DC, American Chemical
breakers additives. The IL modular structure allows engineers to Society, 2002.
tailor properties such as selectivity, viscosity, thermal and chemi- [3] N.V. Plechkova, K.R. Seddon, Chem. Soc. Rev. 37 (2008) 123–150.
[4] J.G. Huddleston, A.E. Visser, W.M. Reichert, H.D. Willauer, G.A. Broker, R.D.
cal stability for specific applications by varying the structure and Rogers, Green Chem. 3 (2001) 156–164.
combination of the constituent ions. Due to the negligible volatility [5] M.J. Earle, J.M.S.S. Esperança, M.A. Gilea, J.N.C. Lopes, L.P.N. Rebelo, J.W.
of the ILs, the azeotrope breaker regeneration for extractive distil- Magee, K.R. Seddon, J.A. Widegren, Nature 439 (2006) 831–834.
[6] M. Smiglak, W.M. Reichert, J.D. Holbrey, J.S. Wilkes, L.Y. Sun, J.S. Thrasher, K.
lation and extraction can be carried out by stripping, evaporation,
Kirichenko, S. Singh, A.R. Katritzky, R.D. Rogers, Chem. Commun. 24 (2006)
drying, or crystallization. Another advantage of ILs as azeotrope 2554–2556.
breakers is the reduction of the effects of conventional solvents [7] M. Petkovic, K.R. Seddon, L.P.N. Rebelo, C.S. Pereira, Chem. Soc. Rev. 40 (2011)
(VOCs) in global climate change. This approach promises excellent 1383–1403.
[8] M. Seiler, C. Jork, A. Kavarnou, W. Arlt, R. Hirsch, AIChE J. 50 (2004) 2439–
contributions to the reduction of greenhouse gases as targeted in 2454.
the most recent United Nations Climate Conference (COP15). [9] W. Arlt, M. Seiler, C. Jork, T. Scheiner, Ionic Liquids as Selective Additives fort
This review has analyzed the contribution of different ILs as sol- he Separation of Close-Boiling or Azeotropic Mixtures, Patent: PCT Int. Appl.
WO 0274718 A2, 2002.
vents for the separation of azeotropes by liquid–liquid extraction [10] V. Julka, M. Chiplunkar, L. O’Young, Chem. Eng. Prog. 105 (2009) 47–53.
and extractive distillation. The increase of the alkyl side chain [11] I. Sucksmith, Chem. Eng. 89 (1982) 91–95.
length of the cation/anion plays a negative role in the ILs’ azeo- [12] S.O. Momoh, Sep. Sci. Technol. 26 (1991) 729–742.
[13] Z. Lei, C. Li, B. Chen, Sep. Purif. Rev. 32 (2003) 121–213.
trope breaking ability. This fact has been demonstrated in many [14] V. Van Hoof, L. Van den Abeele, A. Buekenhoudt, C. Dotremont, R. Leysen, Sep.
cases both in liquid–liquid extraction and extractive distillation Purif. Technol. 37 (2004) 33–49.
and can be taken as a rule of thumb when selecting ILs for the sep- [15] G.W. Meindersma, A.B. de Haan, Chem. Eng. Res. Des. 86 (2008) 745–752.
[16] R.T. Carlin, T.H. Cho, J. Fuller, J. Catalytic immobilized ionic liquid membranes,
aration of azeotropes. The first step in designing any separation in: Proceedings of the Electrochemical Society, Molten Salts XI, 98-11, 1998,
process should take several points into consideration: (i) determi- pp. 180–186.
nation of LLE/VLE data for azeotropic systems + ILs, (ii) evaluation [17] P. Scovazzo, J. Kieft, D.A. Finan, C. Koval, D. DuBois, R. Noble, J. Membrane Sci.
238 (2004) 57–63.
of the separation capacity with decisive parameters such as the
[18] C.A. Cardona, O.J. Sanches, Bioresour. Technol. 98 (2007) 2415–2457.
distribution coefficients and selectivities for liquid–liquid extrac- [19] G.A. Mills, E.E. Ecklund, Ann. Rev. Energy 12 (1987) 47–80.
tion and the relative volatilities and selectivities for extractive dis- [20] R.L. Myers, R.L. Myers, The 100 Most Important Chemical Compounds: A
tillation, and (iii) experimental lab-scale extraction/distillation Reference Guide, Greenwood Press, Westport Conn, 2007. 122.
[21] S.A. Parke, G.G. Birch, Food Chem. 67 (1999) 241–246.
process. [22] H.-J. Huang, S. Ramaswamy, U.W. Tschirner, B.V. Ramarao, Sep. Purif. Technol.
Although only six papers addressed supported liquid mem- 62 (2008) 1–21.
branes for the separation of azeotropes, these data have demon- [23] N. Qureshi, S. Hughes, I.S. Maddox, M.A. Cotta, Bioprocess Biosyst. Eng. 27
(2005) 215–222.
strated that the use of ILs as membrane solvent enables the [24] Dortmund Data Bank, Version 2009, DDBST Software and Separation
supported liquid membrane operation in these processes. The best Technology GmbH, Oldenburg, Germany, 2009.
results are obtained by vapor permeation, which is superior to the [25] I. Wichterle, J. Linek, Z. Wagner, H. V. Kehiaian, Vapor–Liquid Equilibrium in
Mixtures and Solutions, Bibliographic Database, EVLM’2003; Electronic Data
reported values for pervaporation. (ELDATA), Paris, France, 2004.
The systematic analysis of the influence of the structure of the IL [26] J. Zhao, C.C. Dong, C.X. Li, H. Meng, Z.H. Wang, Fluid Phase Equilib. 242 (2006)
on the azeotrope breaking capacity was accomplished in the pres- 147–153.
[27] A.V. Orchilles, P.J. Miguel, E. Vercher, A. Martinez-Andreu, J. Chem. Eng. Data
ent review, with the aim of establishing some guidelines for select- 55 (2010) 1669–1674.
ing the most suitable IL for the separation of specific azeotropic [28] N. Calvar, E. Gomez, B. Gonzalez, A. Dominguez, J. Chem. Eng. Data 55 (2010)
mixtures. Future work should focus on the influence of the IL struc- 2786–2791.
[29] W. Geng, L.Z. Zhang, DS. Deng, Y. Ge, J.B. Ji, J. Chem. Eng. Data 55 (2010) 1679–
ture on the different components of azeotrope systems, in order to
1683.
develop a systematic method for the task-specific tailoring of ILs [30] B. Mokhtarani, J. Gmehling, J. Chem. Thermodyn. 42 (2010) 1036–1038.
for the separation of a azeotropic mixtures. A subsequent goal [31] L.Z. Zhang, Y. Ge, D.X. Ji, J.B. Ji, J. Chem. Eng. Data 54 (2009) 2322–2329.
would be the development of a set of industrial requirements for [32] N. Calvar, B. Gonzalez, E. Gomez, A. Dominguez, J. Chem. Eng. Data 54 (2009)
1004–1008.
ILs characterizing their industrial suitability as entrainers, extrac- [33] N. Calvar, B. Gonzalez, E. Gomez, A. Dominguez, J. Chem. Eng. Data 53 (2008)
tion solvents, absorption agent, and/or membranes. 820–825.
 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28 27

[34] Y. Ge, LZ. Zhang, XC. Yuan, W. Geng, JB. Ji, J. Chem. Thermodyn. 40 (2008) [85] G. Pruckmayr, P. Dreyfuss, M.P Dreyfuss, Polyethers, tetrahydrofuran and
1248–1252. oxetane polymers, in: Kirk–Othmer Encyclopedia of Chemical Technology,
[35] N. Calvar, B. Gonzalez, E. Gomez, A. Dominguez, Fluid Phase Equilib. 259 John Wiley & Sons, Inc., 1996.
(2007) 51–56. [86] Nexant 2004. 1-4. Butanodiol/THF Production, Report published in Nexant
[36] N. Calvar, B. Gonzalez, E. Gomez, A. Dominguez, J. Chem. Eng. Data 51 (2006) on January 23, 2004. <http://nexant.ecnext.com/coms2/gi_0255-14/1-4-
2178–2181. Butanediol-THF.html>.
[37] C. Jork, M. Seiler, Y.-A. Beste, W. Arlt, J. Chem. Eng. Data 49 (2004) 852–857. [87] BASF, Tetrahydrofuran Storage and Handling, BASF Intermediates, Published
[38] X.C. Jiang, J.F. Wang, C.X. Li, L.M. Wang, Z.H. Wang, J. Chem. Thermodyn. 39 on 11.09.2001. <http://basf.com/diols/pdfs/thf_brochure.pdf>.
(2007) 841–846. [88] T. Chang, T.T. Shih, Fluid Phase Equilib. 52 (1989) 161–168.
[39] J.F. Wang, C.X. Li, Z.H. Wang, J. Chem. Eng. Data 52 (2007) 1307–1312. [89] C. Jork, C. Kristen, D. Pieraccini, A. Stark, C. Chiappe, Y.-A. Beste, W. Arlt, J.
[40] J.F. Wang, C.X. Li, Z.H. Wang, Z.J. Li, YB. Jiang, Fluid Phase Equilib. 255 (2007) Chem. Thermodyn. 37 (2005) 537–558.
186–192. [90] X. Hu, J. Yu, H. Liu, Water Sci. Technol. 53 (2006) 245–249.
[41] J. Zhao, C.X. Li, Z.H. Wang, J. Chem. Eng. Data 51 (2006) 1755–1760. [91] E. Sada, T. Morisue, K. Miyahara, J. Chem. Eng. Data 20 (1975) 283–287.
[42] J. Zhao, X.C. Jiang, C.X. Li, Z.H. Wang, Fluid Phase Equilib. 247 (2006) 190–198. [92] M. Seiler, M. Buggert, A. Kavarnou, W. Arlt, J. Chem. Eng. Data 48 (2003) 933–
[43] R.A. Zakariya, J. Zhao, C.X. Li, Z.H. Wang, Chin. J. Chem. Eng. 13 (2005) 791– 937.
795. [93] M. Seiler, Chem. Eng. Technol. 25 (2002) 237–253.
[44] J.F. Wang, X.M. Li, H. Meng, C.X. Li, Z.H. Wang, J. Chem. Thermodyn. 41 (2009) [94] W. Arlt, M. Seiler, G. Sadowski, H. Frey, H. Kautz, Separation of liquids or
167–170. condensable gases, e.g. alcohol from water, involves counter-current
[45] S.A. Chowdhury, J.L. Scott, D.R. MacFarlane, Pure Appl. Chem. 80 (2008) 1325– extractive rectification or liquid extraction with the aid of hyper-branched
1335. polymers as entrainers or solvents, DE Patent 10160518, 2001.
[46] A. Chapeaux, L.D. Simoni, T.S. Ronan, M.A. Stadtherr, J.F. Brennecke, Green [95] D. Bera, D. Lahiri, A. De Leonardis, K.B. De, A. Nag, Agricultural Engineering
Chem. 10 (2008) 1301–1306. International: the CIGR Ejournal 8 (2006) Manuscript FP 06 005.
[47] R.P. Swatloski, A.E. Visser, W.M. Reichert, G.A. Broker, L.M. Farina, R.D. Rogers, [96] L. Berg, Separation of ethyl acetate from ethanol by azeotropic distillation, US
Green Chem. 4 (2002) 81–87. Patent 5993610, 1999.
[48] B.G. Wang, J. Lin, F. Wu, Y. Peng, Ind. Eng. Chem. Res. 47 (2008) 8355–8360. [97] N. Izak, D.B. Van, Separation of ethanol and ethyl acetate from mixtures
[49] A. Arce, J. Martinez-Ageitos, A. Soto, Fluid Phase Equilib. 122 (1996) 117–129. thereof by extractive distillation, South African Patent 2000/4889, 2000.
[50] J.S. Wilkes, J.A. Levisky, R.A. Wilson, C.L. Hussey, Inorg. Chem. 21 (1982) [98] L. Berg, A.-I. Yeh, Separation of methyl acetate from methanol by extractive
1263–1264. distillation, US Patent 4597834, 1986.
[51] K.R. Seddon, A. Stark, M.J. Torres, in: M.A. Abraham, L. Moens, (Eds.), Clean [99] K.A. Horan, C.D. Murphy, R.M. Stephens, J.R. Warner, K.A. Windhorst, Process
SolVents. Alternative Media for Chemical Reactions, ACS Symposyum Series improvement for continuous ethyl acetate production, US Patent 6768021,
819, American Chemical Society, Washington, DC, 2002, pp. 34–49. 2004.
[52] P. Wachter, H.G. Schweiger, F. Wudy, H.J. Gores, J. Chem. Thermodyn. 40 [100] A.V. Orchilles, P.J. Miguel, E. Vercher, A. Martinez-Andreu, J. Chem. Eng. Data
(2008) 1542–1547. 52 (2007) 915–920.
[53] H. Rodriguez, J.F. Brennecke, J. Chem. Eng. Data 51 (2006) 2145–2155. [101] Q.S. Li, J.G. Zhang, Z.G. Lei, J.J. Zhu, F.Y. Xing, J. Chem. Eng. Data 54 (2009) 193–
[54] M. Seiler, W. Arlt, H. Kautz, H. Frey, Fluid Phase Equilib. 201 (2002) 359– 197.
379. [102] Q.S. Li, J.G. Zhang, Z.G. Lei, J.Q. Zhu, J.J. Zhu, X.Q. Huang, Ind. Eng. Chem. Res.
[55] M. Seiler, D. Kohler, W. Arlt, Sep. Purif. Technol. 30 (2003) 179–197. 48 (2009) 9006–9012.
[56] B. Kolbe, J. Gmehling, U. Onken, Ber. Bus. Ges. Phys. Chem. Chem. Phys. 83 [103] A.V. Orchilles, P.J. Miguel, E. Vercher, A. Martinez-Andreu, J. Chem. Eng. Data
(1979) 1133–1136. 52 (2007) 2325–2330.
[57] S.P. Christensen, Fluid Phase Equilib. 150–151 (1998) 763–773. [104] D.L. Zhang, Y.F. Deng, C.B. Li, J. Chen, Ind. Eng. Chem. Res. 47 (2008) 1995–
[58] http://www.inchem.org/documents/ehc/ehc/ehc102.htm#SubSectionNum- 2001.
ber:3.2.2. [105] X.S. Hu, Y.X. Li, D. Cui, B.H. Chen, J. Chem. Eng. Data 53 (2008) 427–433.
[59] http://www.inchem.org/documents/ehc/ehc/ehc103.htm#SubSectionNum- [106] D. Naydenov, H.J. Bart, J. Chem. Eng. Data 52 (2007) 2375–2381.
ber:3.2.2. [107] D. Naydenov, H.J. Bart, J. Chem. Eng. Data 54 (2009) 43–47.
[60] http://www.inchem.org/documents/hsg/hsg/hsg007.htm#SectionNumber: [108] A.B. Pereiro, A. Rodriguez, J. Chem. Thermodyn. 39 (2007) 1608–1613.
1.3. [109] L. Berg, A.-I. Yeh, Separation of methanol from acetone by extractive
[61] P. Dürre, Ann. NY Acad. Sci. 1125 (2008) 353–362. distillation, US Patent 4501645, 1985.
[62] A.V. Orchilles, P.J. Miguel, E. Vercher, A. Martinez-Andreu, J. Chem. Eng. Data [110] L. Berg, Separation of 2-butanone from isopropanol by azeotropic distillation,
53 (2008) 2426–2431. US Patent 5423954, 1995.
[63] L.Z. Zhang, J.Z. Han, R.J. Wang, X.Y. Qiu, J.B. Ji, J. Chem. Eng. Data 52 (2007) [111] L. Berg, Separation of methyl ethyl ketone from ethanol by azeotropic
1401–1407. distillation, US Patent 5868907, 1999.
[64] H.D. Kim, I.C. Hwang, S.J. Park, J. Chem. Eng. Data 55 (2010) 2474–2481. [112] A.V. Orchilles, P.J. Miguel, E. Vercher, A. Martinez-Andreu, J. Chem. Eng. Data
[65] A. Westerholt, V. Liebert, J. Gmehling, Fluid Phase Equilib. 280 (2009) 56– 52 (2007) 141–147.
60. [113] A.B. Pereiro, A. Rodriguez, J. Chem. Eng. Data 52 (2007) 2138–2142.
[66] Q.S. Li, J.G. Zhang, Z.G. Lei, J.Q. Zhu, B.H. Wang, X.Q. Huang, J. Chem. Eng. Data [114] A.B. Pereiro, J. Canosa, A. Rodriguez, Fluid Phase Equilib. 254 (2007) 150–157.
54 (2009) 2785–2788. [115] A. Pucci, Pure Appl. Chem. 61 (1989) 1363–1372.
[67] Q.S. Li, F.Y. Xing, Z.G. Lei, B.H. Wang, Q.L. Chang, J. Chem. Eng. Data 53 (2008) [116] L. Laroche, N. Bekiaris, H.W. Andersen, M. Morari, Can. J. Chem. Eng. 69 (1991)
275–279. 1302–1319.
[68] L.Z. Zhang, D.S. Deng, J.Z. Han, D.X. Ji, J.B. Ji, J. Chem. Eng. Data 52 (2007) 199– [117] S.J. Marwil, Separation of hydrocarbon and alcohol azeotropic mixtures by
205. distillation with anhydrous ammonia, US Patent 4437941, 1984.
[69] L.Z. Zhang, J.Z. Han, D.S. Deng, J.B. Ji, Fluid Phase Equilib. 255 (2007) 179–185. [118] T. Okada, T. Matsuura, in: Proc. Int. Conf. Pervaporation Processes Chem. Ind.
[70] M. Doker, J. Gmehling, Fluid Phase Equilib. 227 (2005) 255–266. 3 (1988) 224-230.
[71] L.Z. Zhang, B.B. Qiao, Y. Ge, D.S. Deng, J.B. Ji, J. Chem. Thermodyn. 41 (2009) [119] M. Laatikainen, M. Lindstrom, Acta Polytech Scand., Chem. Tech. Metall.
138–143. Series 175 (1986) 1–61.
[72] L.D. Simoni, A. Chapeaux, J.F. Brennecke, M.A. Stadtherr, Comput. Chem. Eng. [120] T.M. Letcher, N. Deenadayalu, B. Soko, D. Ramjugernath, P.K. Naicker, J. Chem.
34 (2010) 1406–1412. Eng. Data 48 (2003) 904–907.
[73] X.S. Hu, J. Yu, H.Z. Liu, J. Chem. Eng. Data 51 (2006) 691–695. [121] A.B. Pereiro, F.J. Deive, J.M.S.S. Esperança, A. Rodriguez, Fluid Phase Equilib.
[74] A.G. Fadeev, M.M. Meagher, Chem. Commun. (2001) 295–296. 291 (2010) 13–17.
[75] L.Z. Zhang, X.C. Yuan, B.B. Qiao, R.Z. Qi, J.B. Ji, J. Chem. Eng. Data 53 (2008) [122] U. Domanska, Z. Zolek-Tryznowska, A. Pobudkowska, J. Chem. Eng. Data 54
1595–1601. (2009) 972–976.
[76] L. Berg, Z. Yang, Separation of the propyl alcohols from water by azeotropic or [123] A.B. Pereiro, A. Rodriguez, Green Chem. 11 (2009) 346–350.
extractive distillation, US patent 5085739, 1992. [124] A.B. Pereiro, A. Rodriguez, Chem. Eng. J. 153 (2009) 80–85.
[77] L. Berg, Z. Yang, Separation of tertiary butyl alcohol from water by azeotropic [125] A.B. Pereiro, A. Rodriguez, Fluid Phase Equilib. 270 (2008) 23–29.
or extractive distillation, US patent 5084142, 1992. [126] A.B. Pereiro, A. Rodriguez, Ind. Eng. Chem. Res. 48 (2009) 1579–1585.
[78] E. Vercher, M.I. Vazquez, A. Martinez-Andreu, Fluid Phase Equilib. 202 (2002) [127] A.B. Pereiro, A. Rodriguez, Sep. Purif. Technol. 62 (2008) 733–738.
121–132. [128] A.B. Pereiro, E. Tojo, A. Rodriguez, J. Canosa, J. Tojo, Green Chem. 8 (2006)
[79] E. Vercher, A.V. Orchilles, M.I. Vazquez, A. Martinez-Andreu, Fluid Phase 307–310.
Equilib. 216 (2004) 47–52. [129] A.B. Pereiro, A. Rodriguez, Applications of ionic liquids in azeotropic mixtures
[80] E. Vercher, F.J. Rojo, A. Martinez-Andreu, J. Chem. Eng. Data 44 (1999) 1216– separations, in: A. Kokorin (Ed.), Ionic Liquids, Applications and Perspectives,
1221. INTECH, Rijeka, Croatia, 2011, pp. 225–242. <http://www.intechopen.com/
[81] M.C. Iliuta, F.C. Thyrion, O.M. Landauer, J. Chem. Eng. Data 41 (1996) 402– articles/show/title/applications-of-ionic-liquids-in-azeotropic-mixtures-
408. separations>.
[82] F. Gu, Y. Hou, J. Chem. Eng. Data 45 (2000) 467–470. [130] K.H. Row, Y. Jin, Bioresour. Technol. 97 (2006) 790–793.
[83] J. Ziak, M. Zavodska, J. Jurik, Petrochemia 14 (1974) 41–46. [131] C.L. Schengrund, P. Kovac, J. Lipid Res. 40 (1999) 160–163.
[84] H. Müller, Tetrahydrofuran, in: Ullmann’s Encyclopedia of Industrial [132] M. Stark, H. Jörnvall, J. Johansson, Eur. J. Biochem. 266 (1999) 209–214.
Chemistry, Wiley-VCH, Weinheim, 2002. doi:10.1002/14356007.a26_221. [133] G. Scatchard, C.L. Raymond, J. Am. Chem. Soc. 60 (1938) 1278–1287.
28 A.B. Pereiro et al. / J. Chem. Thermodynamics 46 (2012) 2–28

[134] A.V. Orchilles, P.J. Miguel, E. Vercher, A. Martinez-Andreu, J. Chem. Eng. Data [165] G.W. Meindersma, A. Podt, A.B. de Haan, J. Chem. Eng. Data 51 (2006) 1814–
55 (2010) 1209–1214. 1819.
[135] A.V. Orchilles, P.J. Miguel, E. Vercher, A. Martinez-Andreu, J. Chem. Eng. Data [166] A.L. Revelli, F. Mutelet, J.N. Jaubert, J. Phys. Chem. B 114 (2010) 4600–4608.
53 (2008) 2642–2648. [167] A.B. Pereiro, A. Rodriguez, AICHE J. 56 (2010) 381–386.
[136] G.I. Tatsievskaya, T.A. Vitman, T.M. Kushner, L.A. Serafimov, Zh. Fiz. Khim. 56 [168] J. Zhang, CP. Huang, BH. Chen, PJ. Ren, ZG. Lei, Energy Fuels 21 (2007) 1724–
(1982) 2353–2356. 1730.
[137] S. Wang, Shiyou Huagong 21 (1992) 737–740. [169] N. Deenadayalu, K.C. Ngcongo, T.M. Letcher, D. Ramjugernath, J. Chem. Eng.
[138] N.A. Darwish, Z.A. Al-Anber, Fluid Phase Equilib. 131 (1997) 259–267. Data 51 (2006) 988–991.
[139] F. Whittwer, C. Rose, Dortmund Data Bank V5.0, DDB, Oldenburg, Germany, [170] T.M. Letcher, P. Reddy, J. Chem. Thermodyn. 37 (2005) 415–421.
2007. [171] T.M. Letcher, N. Deenadayalu, J. Chem. Thermodyn. 35 (2003) 67–76.
[140] S. Dernini, R. de Santis, F. Gironi, Ann. Chim. (Rome, Italy) 65 (1975) 409–419. [172] M. Matsumoto, Y. Inomoto, K. Kondo, J. Membr. Sci. 246 (2005) 77–81.
[141] D. Meranda, W.F. Furter, AIChE J. 20 (1974) 103–108. [173] E.J. Gonzalez, I. Dominguez, B. Gonzalez, J. Canosa, Fluid Phase Equilib. 296
[142] K. Weissermel, H.-J. Arpe, Aromatics – production and conversion, in: (2010) 213–218.
Industrial Organic Chemistry, fourth ed., Wiley-VCH, Weinheim, Germany, [174] S.I. Abu-Eishah, A.M. Dowaidar, J. Chem. Eng. Data 53 (2008) 1708–1712.
2003, pp. 313–336. [175] R.J. Wang, C.X. Li, H. Meng, J.F. Wang, Z.H. Wang, J. Chem. Eng. Data 53 (2008)
[143] R. Wennersten, Solvent Extraction Principles and Practice, second ed., Marcel 2170–2174.
Dekker, New York, 2004. [176] R.J. Wang, J.F. Wang, H. Meng, C.X. Li, Z.H. Wang, J. Chem. Eng. Data 53 (2008)
[144] R. Krishna, A.N. Goswami, S.M. Nanoti, B.S. Rawat, M.K. Khanna, J. Dobhal, 1159–1162.
Indian J. Technol. 25 (1987) 602–606. [177] M. Matsumoto, K. Ueba, K. Kondo, Solvent Extr. Res. Dev.-Jpn. 13 (2006) 51–
[145] R.M. De Fré, L.A. Verhoeye, J. Appl. Chem. Biotechnol. 26 (1976) 469–487. 59.
[146] J. Mahmoudi, M.N. Lotfollahi, J. Chem. Thermodyn. 42 (2010) 466–471. [178] M. Matsumoto, K. Ueba, K. Kondo, Desalination 241 (2009) 365–371.
[147] A.S. Al-Jimaz, M.S. Fandary, K.H.A.E. Alkhaldi, J.A. Al-Kandary, M.A. Fahim, Ind. [179] M. Matsumoto, M. Mikami, K. Kondo, J. Jpn. Pet. Inst. 49 (2006) 256–261.
Eng. Chem. Res. 46 (2007) 5686–5696. [180] J. Łachwa, I. Bento, M.T. Duarte, J.N.C. Lopes, L.P.N. Rebelo, Chem. Commun.
[148] A. Cincotti, M. Murru, G. Cao, B. Marongiu, F. Masia, M. Sannia, J. Chem. Eng. (2006) 2445–2447.
Data 44 (1999) 480–483. [181] T.M. Letcher, G.G. Redhi, S.E. Radloff, U. Domanska, J. Chem. Eng. Data 41
[149] Y. Yorulmaz, F. Karpuzcu, Chem. Eng. Res. Des. 63 (1985) 184–190. (1996) 634–638.
[150] W. Wang, Z. Gou, S. Zhu, J. Chem. Eng. Data 43 (1998) 81–83. [182] M.G. Freire, C.M.S.S. Neves, I.M. Marrucho, J.A.P. Coutinho, A.M. Fernandes, J.
[151] S.H. Ali, H.M.S. Lababidi, S.Q. Merchant, M.A. Fahim, Fluid Phase Equilib. 214 Phys. Chem. A 114 (2010) 3744–3749.
(2003) 25–38. [183] V. Najdanovic-Visak, J.M.S.S. Esperanca, L.P.N. Rebelo, M. Nunes da Ponte,
[152] D.F. Schneider, Chem. Eng. Prog. 100 (2004) 34–39. H.J.R. Guedes, K.R. Seddon, J. Szydlowski, Phys. Chem. Chem. Phys. 4 (2002)
[153] G.W. Meindersma, A.R. Hansmeier, A.B. Haan, Ind. Eng. Chem. Res. 49 (2010) 1701–1703.
7530–7540. [184] Y. Ono, Appl. Catal. A: Gen. 155 (1997) 133–166.
[154] A.R. Hansmeier, M. Jongmans, G.W. Meindersma, A.B. de Haan, J. Chem. [185] M.A. Pacheco, C.L. Marshall, Energy Fuels 11 (1997) 2–29.
Thermodyn. 42 (2010) 484–490. [186] K.-Y. Hsu, Y.-C. Hsiao, I.-L. Chien, Ind. Eng. Chem. Res. 49 (2010) 735–749.
[155] J. Garcia, A. Fernandez, J.S. Torrecilla, M. Oliet, F. Rodriguez, J. Chem. Eng. Data [187] L.Y. Wang, J.D. Li, Y.Z. Lin, C.X. Chen, J. Membr. Sci. 305 (2007) 238–246.
55 (2010) 258–261. [188] I. Janisch, H. Landscheidt, W. Struver, A. Klausener, Process for separating of
[156] E. Gomez, I. Dominguez, N. Calvar, A. Dominguez, J. Chem. Thermodyn. 42 methanol from a mixture of dimethyl carbonate and methanol. US patents
(2010) 1234–1239. 351860, 1994.
[157] E.J. Gonzalez, N. Calvar, E. Gomez, A. Dominguez, J. Chem. Eng. Data 55 (2010) [189] A. Arce, H. Rodriguez, A. Soto, Green Chem. 9 (2007) 247–253.
3422–3427. [190] A. Arce, H. Rodriguez, A. Soto, Fluid Phase Equilib. 242 (2006) 164–168.
[158] E.J. Gonzalez, N. Calvar, B. Gonzalez, A. Dominguez, J. Chem. Thermodyn. 41 [191] A. Arce, H. Rodriguez, A. Soto, Chem. Eng. J. 115 (2006) 219–223.
(2009) 1215–1221. [192] A.B. Pereiro, A. Rodriguez, J. Chem. Eng. Data 53 (2008) 1360–1366.
[159] J. Garcia, A. Fernandez, J.S. Torrecilla, M. Oliet, F. Rodriguez, Fluid Phase [193] A.B. Pereiro, A. Rodriguez, J. Chem. Thermodyn. 40 (2008) 1282–1289.
Equilib. 282 (2009) 117–120. [194] V.P. Gupta, E.P. Carey, Extraction of Impurities from Grafted Polyolefins, US
[160] A. Arce, M.J. Earle, H. Rodriguez, K.R. Seddon, A. Soto, Green Chem. 11 (2009) Patent 7060787, 2006.
365–372. [195] I. Rodriguez-Donis, J. Acosta-Esquijarosa, V. Gerbaud, E. Pardillo-Fondevila, X.
[161] A. Arce, M.J. Earle, S.P. Katdare, H. Rodriguez, K.R. Seddon, Phys. Chem. Chem. Joulia, Chem. Eng. Process. 44 (2005) 131–137.
Phys. 10 (2008) 2538–2542. [196] J.G. Wijmans, R.W. Baker, A.P. Mairal, Separation of organic mixtures using
[162] A. Arce, M.J. Earle, H. Rodriguez, K.R. Seddon, Green Chem. 9 (2007) 70–74. gas separation or pervation and dephlegmation, US Patent 6899743, 2005.
[163] A. Arce, M.J. Earle, H. Rodriguez, K.R. Seddon, J. Phys. Chem. B 111 (2007)
4732–4736.
[164] U. Domanska, A. Pobudkowska, M. Krolikowski, Fluid Phase Equilib. 259
(2007) 173–179. JCT 11-184