A review of ionic liquids towards supercritical fluid applications
Seda Keskin, Defne Kayrak-Talay, Ugur Akman ∗, Oner HortacsuDepartment of Chemical Engineering, Bogazici University, Bebek 34342, Istanbul, Turkey
Received 2 August 2006; received in revised form 8 May 2007; accepted 29 May 2007
bstract
Ionic liquids (ILs), considered to be a relatively recent magical chemical due their unique properties, have a large variety of applications inll areas of the chemical industries. The areas of application include electrolyte in batteries, lubricants, plasticizers, solvents and catalysis inynthesis, matrices for mass spectroscopy, solvents to manufacture nano-materials, extraction, gas absorption agents, etc. Non-volatility and non-ammability are their common characteristics giving them an advantageous edge in various applications. This common advantage, when consideredith the possibility of tuning the chemical and physical properties of ILs by changing anion–cation combination is a great opportunity to obtain
ask-specific ILs for a multitude of specific applications. There are numerous studies in the related literature concerning the unique properties,reparation methods, and different applications of ILs in the literature. In this review, a general description of ILs and historical background areiven; basic properties of ILs such as solvent properties, polarity, toxicology, air and moisture stability are discussed; structure of ILs, cation, anionypes and synthesis methods in the related literature are briefly summarized. However, the main focus of this paper is how ILs may be used in thehemicals processing industries. Thus, the main application areas are searched and the basic applications such as solvent replacement, purificationf gases, homogenous and heterogeneous catalysis, biological reactions media and removal of metal ions are discussed in detail. Not only thedvantages of ILs but also the essential challenges and potentials for using ILs in the chemical industries are also addressed. ILs have becomehe partner of scCO2 in many applications and most of the reported studies in the literature focus on the interaction of these two green solvents,.e. ILs and scCO . The chemistry of the ILs has been reviewed in numerous papers earlier. Therefore, the major purpose of this review paper is
2
o provide an overview for the specific chemical and physical properties of ILs and to investigate IL–scCO2 systems in some detail. Recovery ofolutes from ILs with CO2, separation of ILs from organic solvents by CO2, high-pressure phase behavior of IL–scCO2 systems, solubility of ILsn CO2 phase, and the interaction of the IL–scCO2 system at molecular level are also included.
Ionic liquids (ILs) have been accepted as a new green chemi-al revolution which excited both the academia and the chemicalndustries. This new chemical group can reduce the use ofazardous and polluting organic solvents due to their uniqueharacteristics as well as taking part in various new syntheses.he terms room temperature ionic liquid (RTIL), nonaqueous
onic liquid, molten salt, liquid organic salt and fused salt havell been used to describe these salts in the liquid phase [1]. ILsre known as salts that are liquid at room temperature in con-rast to high-temperature molten salts. They have a unique arrayf physico-chemical properties which make them suitable inumerous applications in which conventional organic solventsre not sufficiently effective or not applicable. Short [2] pointedut in 1980, that there were only a few patent applications forLs, in 2000, the number of patent applications increased to 100,nd finally by 2004, there were more than 800. This is a clearndication of the high affinity of the academia and industry tohe ILs.
. History of ILs
ILs have been known for a long time, but their extensive uses solvents in chemical processes for synthesis and catalysis hasecently become significant. Welton [1] reported that ILs areot new, and some of the ILs such as [EtNH3][NO3] was firstescribed in 1914 [3]. The earliest IL in the literature was createdntentionally in 1970s for nuclear warheads batteries [4]. During940s, aluminum chloride-based molten salts were utilized forlectroplating at temperatures of hundreds of degrees Celsius.n the early 1970s, Wilkes tried to develop better batteries foruclear warheads and space probes which required molten saltso operate [4]. These molten salts were hot enough to damagehe nearby materials. Therefore, the chemists searched for salts
hich remain liquid at lower temperatures and eventually they
dentified one which is liquid at room temperature. Wilkes andis colleagues continued to improve their ILs for use as batterylectrolytes and then a small community of researchers began
o make ILs and test their properties [5,6]. In the late 1990s, ILsecame one of the most promising chemicals as solvents.
The first ILs, such as organo-aluminate ILs, have lim-ted range of applications because they were unstableo air and water. Furthermore, these ILs were not inertowards various organic compounds [7]. After the firsteports on the synthesis and applications of air stableLs such as 1-n-butyl-3-methlyimidazolium tetrafluoroborate[bmim][BF4]) and 1-n-butyl-3-methlyimidazolium hexafluo-ophosphate ([bmim][PF6]), the number of air and water stableLs has started to increase rapidly [7]. Recently, researchers haveiscovered that ILs are more than just green solvents and theyave found several applications such as replacing them witholatile organic solvents, making new materials, conducting heatffectively, supporting enzyme-catalyzed reactions, hosting aariety of catalysts, purification of gases, homogenous and het-rogeneous catalysis, biological reactions media and removal ofetal ions [4].Some of the basic physical properties of ILs such as density
nd viscosity are still being evaluated by the researchers sincehe study of the IL is a relatively young field [8]. The numberf research on ILs and their specific applications is increasingapidly in the literature. For example, the cation 1-n-ethyl-3-ethylimidazolium has been the most widely studied until 2001,
nd nowadays, 1-3-dialkyl imidazolium salts are the most popu-arly used and investigated class of ILs. For the future of ILs, theim of research is the commercialization of ILs in order to usehem as solvents, reagents, catalysts and materials in large-scalehemical applications.
. Basic properties of ILs
ILs are made of positively and negatively charged ions,hereas water and organic solvents, such as toluene andichloromethane, are made of molecules. The structure of ILs is
imilar to the table salt such as sodium chloride which containsrystals made of positive sodium ions and negative chlorine ions,ot molecules. While, salts do not melt below 800 ◦C, most of ILsemain liquid at room temperature. The melting points of sodium
1 ritical Fluids 43 (2007) 150–180
crmlhpas
p[sTpahces
o[
•
•••••
•
••
•
mIcbf
iasi
aFti
o
vasa(vtwep
oTctoospocaw
gaiNtIt
itteia[18,19].
52 S. Keskin et al. / J. of Superc
hloride and lithium chloride are known as 801 and 614 ◦C,espectively. Since these conventional molten salts exhibit highelting points, their use as solvents in applications is severely
imited. However, RTILs are liquid generally up to 200 ◦C. ILsave a wide liquidus ranges. The adopted upper melting tem-erature limit for the classification as ‘IL’ is known as 100 ◦Cnd higher melting ion systems are generally referred as moltenalts.
Researchers explained that ILs remain liquid at room tem-erature due to the reason that their ions do not pack well9]. Combination of bulky and asymmetrical cations and evenlyhaped anions form a regular structure namely a liquid phase.he low melting points of ILs are a result of the chemical com-osition. The combination of larger asymmetric organic cationnd smaller inorganic counterparts lower the lattice energy andence the melting point of the resulting ionic medium. In someases, even the anions are relatively large and play a role in low-ring the melting point [10]. Most widely used ILs and theirtructures are given in Table 1.
As solvents, ILs posses several advantages over conventionalrganic solvents, which make them environmentally compatible1,4,8,10–15]:
ILs have the ability to dissolve many different organic, inor-ganic and organometallic materials.ILs are highly polar.ILs consist of loosely coordinating bulky ions.ILs do not evaporate since they have very low vapor pressures.ILs are thermally stable, approximately up to 300 ◦C.Most of ILs have a liquid window of up to 200 ◦C whichenables wide kinetic control.ILs have high thermal conductivity and a large electrochem-ical window.ILs are immiscible with many organic solvents.ILs are nonaqueous polar alternatives for phase transfer pro-cesses.The solvent properties of ILs can be tuned for a specificapplication by varying the anion cation combinations.
Generally, the above statements are valid for the most com-only used ILs. However, one should note that there are many
Ls containing different anions and cations and their propertiesover a vast range. Therefore, the above statements should note generalized for all existing ILs and for those designed in theuture.
ILs exhibit the ability to dissolve a wide variety of materialsncluding salts, fats, proteins, amino acids, surfactants, sugarsnd polysaccharides. ILs have very powerful solvent propertiesuch that they can dissolve a wide range of organic molecules,ncluding crude oil, inks, plastics, and even DNA [9].
Two important groups of ILs are those based on imidazoliumnd pyridinium cations with PF6
− and BF4− anions [13,14].
igs. 1 and 2 illustrate the imidazolium and pyridinium deriva-
ives of ILs and their possible anions which are extensivelynvestigated in literature.
ILs tend not to give off vapors in contrast to traditionalrganic solvents such as benzene, acetone, and toluene. The
Fig. 1. Imidazolium derivatives of ILs (www.sigmaaldrich.com).
apor pressures of the ILs are extremely low and are considereds negligible. For example, Kabo et al. [16] gave the vapor pres-ure of [bmim][PF6] at 298.15 K as 10−11 Pa. ILs are introduceds green solvents because unlike the volatile organic compoundsVOCs) they replace, many of these compounds have negligibleapor pressure, they are not explosive and it may be feasibleo recycle and repeatedly reuse them. It is more convenient toork with ILs in the laboratory since the non-evaporating prop-
rties of ILs eliminate the hazardous exposure and air pollutionroblems.
ILs are also known as ‘designer solvents’ since they give thepportunity to tune their specific properties for a particular need.he researchers can design a specific IL by choosing negativelyharged small anions and positively charged large cations, andhese specific ILs may be utilized to dissolve a certain chemicalr to extract a certain material from a solution. The fine-tuningf the structure provides tailor-designed properties to satisfy thepecific application requirements. The physical and chemicalroperties of ILs are varied by changing the alkyl chain lengthn the cation and the anion. For example, Huddleston et al. [17]oncluded that density of ILs increases with a decrease in thelkyl chain length on the cation and an increase in the moleculareight of the anion.Although ILs are studied by a great number of research
roups, there are still many questions that scientist are notble to answer. For example, one of the basic rules of chem-stry “like dissolves like” is seem to be broken by some ILs:onpolar benzene is up to 50% soluble (by volume) in polar
etrachloroaluminate-based ILs [9]. Therefore, studies on whyLs are able to dissolve uncharged covalent molecules are con-inuing.
Until recently, ILs have been considered to be scarce but its now known that many salts form liquids at or close to roomemperature. There are literally billions of different structureshat may form an IL. The composition and the specific prop-rties of these liquids depend on the type of cation and anionn the IL structure. By combining various kinds of cation andnion structures, it is estimated that 1018 ILs can be designed
Fig. 2. Pyridinium derivatives of ILs (www.sigmaaldrich.com).
eprinted with permission from [121]. Copyright 2004 American Chemical So
.1. Solvent properties of ILs
Both the chemical industry and academia search for alter-ative solvents to meet the cleaner technology requirements
sioa
ince the most widely used solvents are volatile and damag-ng. ILs are good solvents for a wide range of substances;rganic, inorganic, organometallic compounds, bio-moleculesnd metal ions. They are usually composed of poorly coordinat-
1 ritica
isvtcwonvwtoip
vbrcuhmndTcccs[Idladatci
aepiptua
aaaatpsp
3
vrpbUit
btsst1[2e[teocw
3
nbmAttactotas
egDtsobvt
54 S. Keskin et al. / J. of Superc
ng ions which makes them highly polar but non-coordinatingolvents. ILs are immiscible with most of the organic sol-ents, thus they provide a nonaqueous, polar alternative forwo-phase systems [19]. Furthermore, ILs which are not mis-ible with water can be used as immiscible polar phasesith water. Although all other conventional solvents evap-rate to the atmosphere, ILs do not evaporate and theironvolatility gives an opportunity to utilize them in high-acuum systems. The negligible volatility is the basic propertyhich characterizes them as green solvents. Considering poten-
ial as solvents, ILs can easily replace other conventionalrganic solvents which are used in large quantities in chem-cals processing industries to eliminate major environmentalroblems.
Many chemical reactions are carried out in conventional sol-ents. Upon the completion of reaction, chemical products muste taken out of the solvent. There are several techniques toecover a product from a solvent: For example, water-solubleompounds may be extracted with water; distillation may besed for chemicals with high vapor pressures. On the otherand, for the chemicals with low vapor pressures, distillationust be performed at low pressures, which may not be eco-
omical. In addition to this, there are some chemicals that canecompose as a result of heating, such as pharmaceuticals.herefore, ILs seem to be potentially good solvents for manyhemical reactions in the cases where distillation is not practi-al, or water insoluble or thermally sensitive products are theomponents of a chemical reaction. Although, ILs are not con-idered to be distilled due to their low volatility, Earle et al.20] showed that many ionic liquids, especially bistriflamideLs, can be distilled at 200–300 ◦C and low pressure withoutecomposition. It was once more understood that there is aong way for total investigation of the properties of ILs. Theuthors suggested that the possibility of IL distillation intro-uced a new method for IL purification, and also new applicationreas (such as isolation of highly soluble products by high-emperature crystallization) could emerge. But, distillation stillannot be applied when heat-labile products are encounteredn ILs.
In most chemical applications, extraction is used for sep-ration since it is an energy efficient technique. Generally,xtraction consists of two immiscible phases such as an organichase and an aqueous phase. Many organic solvents usedn extractions are known with their flammable and toxicalroperties. In order to improve the safety and environmen-al friendliness of this conventional technique, ILs may besed as ideal substitutes due to their stability, nonvolatility anddjustable miscibility and polarity [15].
The solvent properties of ILs are mainly determined by thebility of the salt to act as a hydrogen bond donor and/orcceptor and the degree of localization of the charges on thenions [1,21]. Charge distribution on the anions, H-bondingbility, polarity, dispersive interactions are the main factors
hat influence the physical properties of ILs [22]. For exam-le, imidazolium-based ILs are highly ordered hydrogen-bondedolvents and they have strong effects on chemical reactions androcesses.
t
sp
l Fluids 43 (2007) 150–180
.2. Polarity of ILs
Polarity of chemicals is commonly used to classify the sol-ents. The terms used as polar, nonpolar and apolar are generallyelated to the values of dielectric constants, dipole moments,olarizabilities. If a solvents has the ability to dissolve and sta-ilize dipolar or charged solutes, it is defined as a polar solvent.nder this simple definition, ILs are highly polar solvents, but
t is not completely true to make such strict conclusions sincehere ILs can be designed in a vast range.
The existences of polar and nonpolar domains, believed toe associated with the unique “amphiphilic” solvent proper-ies of ILs, are found in the structures of PF6
− and BF4−
alts [23]. Since polarity is the simplest indicator of solventtrength, researchers compared polarities of ILs and conven-ional solvents: Carmichael and Seddon [24] showed that-alkyl-3-methylimidazolium ILs with anions [PF6], [BF4],(CF3SO2)2N], and [NO3] are in the same polarity region as-aminoethanol and lower than alcohols such as methanol,thanol and butanol. Aki et al. [25] indicated that [bmim][PF6],C8mim][PF6], [bmim][NO3] and [N-bupy][BF4] are more polarhan acetonitrile and less polar than methanol and these ILs arexpected to be at least partially miscible with water. ILs basedn [PF6] anion is preferred as solvents in most extraction appli-ation to form biphasic systems due to their immiscibility withater.
.3. Toxicology of ILs
The green character of ILs has been usually related with theiregligible vapor pressure; however their toxicology data haveeen very limited until now. Several authors [26–29] alreadyentioned this lack of toxicological data in the literature [30].lthough ILs will not evaporate and thus will not cause air pollu-
ion, it does not mean that they will not harm the environment ifhey enter. Most of ILs are water soluble and they may enter thequatic environment by accidental spills or effluents. The mostommonly used ILs [bmim][PF6] and [bmim][BF4] are knowno decompose in the presence of water and as a result hydroflu-ric and phosphoric acids are formed [31]. Therefore, bothoxicity and ecotoxicity information which provide metabolismnd degradability of ILs are also required to label them as greenolvents or investigate their environmental impact.
The ecotoxicological studies performed to understand theffects of different ILs on enzymatic activities, cells and microor-anisms are utilized to obtain LC50 levels (lethal concentration).ecreasing LC50 values indicate higher toxicities according to
he toxicity classes of Hodge and Sterner scale (1956) [32]. Thiscale indicates that the LC50 value (in terms of mg/L) of 10r less shows that the chemical is extremely toxic, LC50 valueetween 10 and 100 shows that chemical is highly toxic, LC50alue between 100 and 1000 shows that chemical is slightlyoxic, and finally LC50 value between 1000 and 10,000 means
hat chemical is practically nontoxic.
The impact of ILs on aquatic ecosystems is highly importantince some of ILs have a high solubility in water. Maginn [33]rovided the LC50 levels for two imidazolium-based ILs with
aphnia magna, common fresh water crustaceans. Due to theeason that D. magna are filter feeders at the base of the aquaticood chain, their responses to ILs are essential to understand howhese new solvents may impact an environmental ecosystem. AsLs, 1-n-butyl 3-methylimidazolium cation with PF6
− and BF6−
nions are used and the results are tabulated in Table 2: Thesewo ILs are as toxic to Daphnia as benzene and even far moreoxic than acetone, but much less toxic than ammonia, chlorine,henol, etc. Wells and Coombe [34] also provided the resultsf freshwater ecotoxicity tests of some common ILs with imi-azolium, ammonium, phosphonium and pyridinium cations onnvertebrate D. magna and the green alga Pseudokirchneriellaubcapitata (formerly known as Selenastrumcapricornutum).he results were reported using medium effective concentra-
ion (EC50) values. The toxicity values of the most toxic ILere four orders of magnitude more than the least toxic IL.here was a relation between the order of toxicity and alkylide chain length of the cation. For alkyl methylimidazoliumLs with C4 side chain constituents showed moderate toxicity,hereas the C12, C16, and C18 species were very highly toxic
o both organisms under investigation. Pyridinium, phospho-ium, and ammonium species with C4 side chain constituentsad also only moderate toxicity, whereas C6 and longer sidehains showed significant increases in toxicity. It was shownhat the least toxic ionic liquids’ ecotoxicity were comparable toydrocarbons’ such as toluene and xylene. The most toxic ioniciquids are many orders of magnitude more acutely ecotoxichan organic solvents such as methanol, tert-butyl methyl ether,cetonitrile, and dichloromethane. The authors also emphasizedhat simple acute ecotoxicity measurements did not enough toully characterize the full impact of a solvent released to thenvironment but were only part of the environmental impactssessment.
There were also some studies on investigation of toxicity ofLs performed on animals such as the nematode model organ-sm (Caenorhabditis elegans) [35], freshwater pulmonate snailsPhysa acuta) [36], Fischer 344 rats [37] and zebra fish (Danioerio) [38].
One of the most important points that must be taken intoccount during the toxicological study of the ILs is to pay atten-
ion to the purity of the IL studied. Therefore, different authors35,39] attached significant importance to proper analyzing tech-iques [30]. ILs are introduced under the concept of greenhemistry in all research papers due to their nonvolatile nature.
cotd
l Fluids 43 (2007) 150–180 155
he environmental persistence [34,40] of commonly used ILs,long with their possible toxicity should be taken into account.here are still very few results about the (eco)toxicological effectf ILs and they can be evaluated more satisfactorily as greenolvents or not after more data on the subject will be provided.he possible toxic and non-biodegradable nature of the existing
Ls also led to the development of new types of nontoxic andiodegradable ILs [40–46].
.4. Air and moisture stability of ILs
The stability of ILs is crucial for optimum performance. Manyf ILs are both air and moisture stable, some are even hydropho-ic. On the other hand, most imidazolium and ammonium saltsre hydrophilic and if they are used in open vessels, hydrationill certainly occur. The hydrophobicity of an IL increases with
ncreasing length of the alkyl chain [25]. Despite their widepread usage, ILs containing PF6
− and BF4− have been reported
o decompose in the presence of water, giving off HF. Wasser-cheid et al. [47] pointed out that ILs containing halogen anionsenerally show poor stability in water, and also give off toxic andorrosive species such as HF or HCl. Therefore, they suggest these of halogen-free and relatively hydrolysis-stable anions suchs octylsulfate-compounds.
The degree to which this hydration is a problem dependsn the application. For instance, small amounts of highly reac-ive species which are used as catalysts may be deactivated byven very small amounts of water. For this kind of application,Ls must be handled under an inert atmosphere. Moreover, theolutes used may be sensitive for air or moisture, thus an inerttmosphere is required for the IL–solute systems.
The interaction between water and ILs and their degree ofydroscopic character are strongly dependent on anions. Themount of absorbed water is highest in the BF4
− and lowest inF6
− [48]. However, Tf2N− is much more stable in the pres-nce of water as well as having the advantage of an increasedydrophobic character.
ILs immiscible with water tend to absorb water from thetmosphere. The infra-red (IR) studies of Cammarata et al. [31]emonstrated that the water molecules absorbed from the airre mostly present in the free state, bonded via H-bonding withhe PF6
− and BF4− anions. The presence of water may have
ramatic effect on IL reactivity. Since water is present in all ILs,hey are usually utilized after a moderate drying process.
The new ILs synthesized are more stable than the oldalogenoaluminate systems. Certain ILs incorporating 1-3-ialkyl imidazolium cations are generally more resistant thanraditional solvents under harsh process conditions, such as thoseccurring in oxidation, photolysis and radiation processes [10].
. Structure and synthesis of ILs
There are a great number of different cation and anion
ombinations to synthesize IL. Different types of ILs give anpportunity to modify the physical and chemical properties ofhe IL. The most widely used cations are imidazolium, pyri-inium, phosphonium and ammonium. The properties of ILs
156 S. Keskin et al. / J. of Supercritica
F
aespptiwhoa
4
ibdsn
maicSodwiasahm
uaptouba
cbfmbw(
4
lpdrromc
osoapaiaw
4
sMcaztpwuntptiohbi
ig. 3. Most commonly used cation structures and possible anion types [50].
re determined by mutual fit of cation and anion, size, geom-try, and charge distribution. Among the similar class of salts,mall changes in types of ions influence the physico-chemicalroperties. The overall properties of ILs result from the com-osite properties of the cations and anions and include thosehat are superacidic, basic, hydrophilic, water miscible, watermmiscible and hydrophobic. Usually, the anion controls theater miscibility, but the cation also has an influence on theydrophobicity or hydrogen bonding ability [49]. The structuresf most commonly used cations and some possible anion typesre tabulated in Fig. 3 [50].
.1. Anions
The properties of ILs are determined by the anion type. Thentroduction of different anions results in an increasing num-er of alternative ILs with various properties. There are twoifferent types of IL anions: ILs containing fluorous anionsuch as PF6
−, BF4−, CF3SO3
−, (CF3SO3)2N− and ILs withon-fluorous anions such as AlCl4−.
In designing ILs, fluorous anions are usually used. Theost popular anions consist of chloride, nitrate, acetate, hex-
fluorophosphate and tetrafluoroborate [13]. The most widelynvestigated ILs are the ones with anions PF6
− and BF4−. Espe-
ially PF6− is the most prominent anion used in IL research.
ince the anion chemistry has a large effect on the propertiesf IL, although the cations are the same, there are significantifferences between ILs with different anions. For example ILith 1-n-butyl-3-methylimidazolium cation and PF6
− anion ismmiscible with water, whereas IL with same cation and BF4
−nion is water soluble. This example represents the ‘designerolvent’ property of ILs: different ion pairs determine physicalnd chemical properties of the liquid. By changing the anion theydrophobicity, viscosity, density and solvation of the IL systemay be changed [8].Although PF6
− and BF4− are the two anion types that are
tilized in most of IL applications, they have an important dis-dvantage: these two anions may decompose when heated in theresence of water and liberate HF. After the researchers realizedhe production of HF in the presence of water, the bonding style
f anion was altered and fluorous anions inert to hydrolysis weresed. The fluorine of the anion is bonded to carbon and C–F bondecomes inert to hydrolysis. In this way, ILs such as CF3SO3
−nd (CF3SO3)2N− are produced [50].
5
p
l Fluids 43 (2007) 150–180
Fluorinated anions tend to be expensive and in response toost and safety concerns new ILs with non-fluorous ions haveeen introduced. In the synthesis of these ILs, anions are derivedrom inexpensive bulk chemicals. Alkylsulfate anions are theost popular non-fluorous anions due to their nontoxic and
iodegradable structures. The first commercially available IL forhich toxicology data are available contains alkylsulfate anion
methosulfate) [50].
.2. Cations
The cation of IL is generally a bulk organic structure withow symmetry. Most ILs are based on ammonium, sulfonium,hosphonium, imidazolium, pyridinium, picolinium, pyrroli-inium, thiazolium, oxazolium and pyrazolium cations. Theesearch mainly focuses on RTILs composed of asymmet-ic N,N-dialkylimidazolium cations associated with a varietyf anions. 1-n-butyl-3-methylimidazolium and 1-n-ethyl-3-ethylimidazolium are the most investigated structures of this
lass.Chiappe and Pieraccini [18] indicated that the melting points
f the most ILs are uncertain since ILs undergo considerableupercooling. Therefore, by examining the properties of a seriesf imidazolium cation based ILs, it has been concluded thats the size and asymmetry of the cation increases, the meltingoint decreases. Further, an increase in the branching on thelkyl chain increases the melting point. The melting point of ILss essential because it represents the lower limit of the liquiditynd with thermal stability it defines the interval of temperaturesithin which it is possible to use ILs as solvents [18].
.3. Synthesis
There are three basic methods to synthesize ILs: metathe-is reactions, acid–base neutralization, direct combination [1].
any alkylammonium halides are commercially available; theyan also be prepared simply by the metathesis reaction of theppropriate halogenoalkane and amine. Pyridinium and imida-olium halides are also synthesized by metathesis reaction. Onhe other hand, monoalkyllammonium nitrate salts are best pre-ared by the neutralization of aqueous solutions of the amineith nitric acid. After neutralization reactions, ILs are processednder vacuum to remove the excess water [1]. Tetraalkylammo-ium sulfonates are also prepared by mixing sufonic acid andetraalkylammonium hydroxide [51]. In order to obtain pure IL,roducts are dissolved in an organic solvent such as acetoni-rile and treated with activated carbon, and the organic solvents removed under vacuum. The final method for the synthesisf ILs is the direct combination of halide salt with a metalalide. Halogenoaluminate and chlorocuprate ILs are preparedy this method. The synthesis methods of ILs have been givenn numerous articles [52–56].
. Major applications suggested for ILs
The research areas on ILs are growing very rapidly and theotential application areas of ILs are numerous. The unique
S. Keskin et al. / J. of Supercritica
caaocffamcaa
5
amAsiausmaida(iscaa
asTh
ihp[wzo
umttaIwb
uaaelAbrr
5
nidbtufa
ga[seaitiwtwoaCellowlw
Fig. 4. Major application areas of ILs.
hemical and physical properties of ILs bring about severalpplication areas including reaction and synthesis media. Thepplication areas of ILs can be expressed as solvents for organic,rganometallic synthesis and catalysis; electrolytes in electro-hemistry, in fuel and solar cells; lubricants; stationary phasesor chromatography; matrices for mass spectrometry; supportsor the immobilization of enzymes; in separation technologies;s liquid crystals; templates for synthesis nano-materials andaterials for tissue preservation; in preparation of polymer–gel
atalytic membranes; in generation of high conductivity materi-ls [7]. Fig. 4 represents the major applications suggested for ILsnd these essential applications are discussed in detail below.
.1. Solvent replacement
A majority of common solvents have potential health hazardslthough they are extensively utilized. For example, approxi-ately half of 189 hazardous air pollutants regulated by Cleanir Act Amendment of U.S. (1990) are VOCs including solvents
uch as dichloromethane [57]. The VOCs are the workhorses ofndustrial chemistry in the pharmaceutical and petrochemicalreas. The use of VOCs by these industries can be assessedsing the Sheldon E-factor. This factor is responsible to mea-ure process by-products as a proportion of production on theass basis. Researchers investigated how VOC use is distributed
cross the chemical industry and found that the value of E-factors between 25 and 100 for pharmaceuticals industries with a pro-uction of 10 to 103 t/year although oil refining industries withproduction of 106 to 108 t/year have an E-factor of 0.1 [9]
adapted from [58]). These values suggest that pharmaceuticalsndustries use inefficient and dirty processes although on smallercale as compared to the oil refining industries. The oil and bulkhemicals industries which are commonly regarded as dirty arepparently remarkably waste conscious when the E-factors arenalyzed [9].
As the introduction of cleaner technologies has becomemajor concern throughout both industry and academia, the
earch for the alternative solvents has become a high priority.herefore, environmentally friendly ILs can easily replace theazardous VOCs in large scale to reduce E-factors.
ILs are able to dissolve a variety of solutes. They can be usednstead of traditional solvents in liquid–liquid extractions whereydrophobic molecules such as simple benzene derivatives willartition to the IL phase. Huddleston et al. [17] showed that
bmim][PF6] could be used to extract aromatic compounds fromater. Fadeev and Meagher [59] demonstrated that two imida-
olium ILs with PF6− anion could be used for the extraction
f butanol from aqueous fermentation broths. Selvan et al. [60]
saCi
l Fluids 43 (2007) 150–180 157
sed ILs for the extraction of aromatics from aromatic/alkaneixtures, whereas Letcher et al. [61] used ILs for the extrac-
ion of alcohols from alcohol/alkane mixtures. Moreover, binaryemperature–composition curves of ILs with alcohols, alkanes,romatics and water; ternary temperature–composition curves ofLs with alcohols and water; solubilities of some organics andater in ILs are all investigated by various groups to completelyenefit from the solvent properties of ILs [62–64].
Arce et al. [65] studied essential oil terpenless by extractionsing organic solvents or ILs. Citrus essential oil is simulateds a mixture of limonene and linalool and 2-butene-1,4-diolnd ethylene glycol are used as solvents. They choose 1-thyl-3-methylimidazolium methanesulfonate as the IL andiquid–liquid equilibria data for the ternary systems are reported.rce et al. [65] concluded that IL presents the highest selectivityut close to the other organic solvents and they reported that theesults for solute distribution ratio depend on the concentrationange of extraction.
.2. Purification of gases
Reliable information on the solubility of gases in ILs iseeded for the design and operation of any possible processesnvolving IL. Processes using ILs to purify gas streams wereeveloped after solubilities of various gases in ILs were reportedy some researchers [66–68]. These experimental studies showhat some gases, especially CO2 is highly soluble in ILs. The sim-lations performed explain that the anion of the IL is responsibleor high gas solubility. With this property ILs, can be replaceds solvents in reactions involving gaseous species.
Anthony et al. [69] investigated solubility of nine differentases up to 13 bar: carbon dioxide, ethylene, ethane, methane,rgon, oxygen, carbon monoxide, hydrogen, and nitrogen inbmim][PF6]. These gases were chosen for several reasons: CO2olubility is important due to the possibility of using scCO2 toxtract solutes from ILs; ethylene, hydrogen, carbon monoxide,nd oxygen are reactants in several types of reactions studiedn IL such as hydroformylations, hydrogenations, and oxida-ions. Due to the nonvolatile nature of IL, the gas solubilitiesn IL were measured using a gravimetric technique, usuallyith a microbalance. The study of Anthony et al. [69] showed
hat CO2 has the highest solubility and strongest interactionsith [bmim][PF6], followed by ethylene and ethane. Argon andxygen had very low solubilities and immeasurably weak inter-ctions. Fig. 5 demonstrates the solubility of various gases (CO2,2H4, C2H6, CH4, Ar, O2) in [bmim][PF6] at 25 ◦C and at differ-nt pressures. Except for CO2, all gases remained in the Henry’saw regime up to 13 bar. However, CO2 showed some non-inearity, indicating some degree of saturation. Henry’s constantsf these gases in various organic solvents and in [bmim][PF6]ere compared and the results showed that the gases that are
ess soluble in the IL are less soluble in the other solvents asell. However, CO2 is more soluble in the IL than in the other
olvents. The relatively high solubility of CO2 was explaineds a result of its large quadrapole moment. The solubility ofO2 in [bmim][PF6] at different temperatures is demonstrated
n Fig. 6.
158 S. Keskin et al. / J. of Supercritica
Fp
Cm(htsoRetatettae
wso
Fw
ttTgadcaatombdIl
c(mbsetapd
5
tnito
ig. 5. Solubility of various gases in [bmim][PF6] at 25 ◦C. (Reprinted withermission from [69]. Copyright 2002 American Chemical Society)
Camper et al. [66] measured the solubility of CO2 and2H4 in [bmim][PF6], [emim][Tf2N], [emim][CF3SO3] (ethyl-ethylimidazolium trifluoromethanesulfone), [emim][dca]
ethylmethylimidazolium dicyanamide) and [thtdp][Cl] (tri-exyltetradecylphosphonium chloride) to demonstrate thathe regular solution theory can be used to model the gasolubilities in RTILs at low pressures and studied the effectsn pressure and the temperature on the solubility of gases inTILs. The previous works; Blanchard et al. [70] and Anthonyt al. [71] related that the solubility of the gases in ILs tohe intermolecular interactions between the anion of the ILnd the gas. On the other hand, Camper et al. [66] indicatedhat at low pressures, the solubility of CO2 and C2H4 may bexplained using the regular solution theory without consideringhe intermolecular interactions between the anion of the IL andhe gas. At higher pressures, regular solution theory is limitednd Camper et al. [66] attributed this limitation to the dominantntropic effects.
Recently, the results of solubility of hydrogen in [bmim][PF6]
as presented for temperatures from 313 to 373 K and pres-
ures up to 9 MPa. The results demonstrated that the solubilityf hydrogen in [bmim][PF6] is low and increases slightly with
ig. 6. Solubility of CO2 in [bmim][PF6] at different temperatures. (Reprintedith permission from [69]. Copyright 2002 American Chemical Society)
awcmsuMois
Salh[cnrfl
l Fluids 43 (2007) 150–180
emperature [72]. Since ILs can dissolve certain gaseous species,hey may be used in conventional gas absorption applications.he nonvolatility of ILs prevent any cross contamination of theas stream by the solvent during the process. Moreover, regener-tion of the solvent may be performed easily by a simple flash oristillation to remove the gas from the solvent without any crossontamination. The other advantages of ILs as separating agentsre no solvent loss and no air pollution. Currently, researchersre interested in examining the potential of ILs for the separa-ion of CO2 from flue gases emitted from fossil–fuel combustionperations [73]. ILs may also be utilized as supported liquidembranes. In conventional membranes, gas dissolves in liquid
ut then the liquid in which the gas dissolved evaporates ren-ering the membrane useless [50]. Due to the nonvolatility ofLs, they can be immobilized on a support and used in supportediquid membranes.
ILs are also used for storage and delivery of hazardous spe-ialty gases such as phosphine (PH3), arsine (AsH3) and stibineSbH3). GASGUARD® Sub-Atmospheric Systems supply theajor ion implant gases: AsH3, boron trifluoride (BF3), enrich
oron trifluoride (11BF3) and PH3 sub-atmospherically [74]. Theystem is combined with gas supply technologies for the deliv-ry of the gases when needed. In the complexed gas technology,he desired gases (BF3 and PH3) are chemically bond to ILs sub-tmospherically, then pulling the vacuum on the IL–gas complexrovides the mechanism to evolve high purity gas, similar toesorbing a gas from active carbon.
.3. Homogenous and heterogeneous catalysis
One of the most important targets of modern chemistry iso combine the advantages of both homogenous and heteroge-eous catalysis [75]. Greater selectivity is generally observedn homogenous catalysis compared to its heterogeneous coun-erparts, but separation of the catalyst from the product streamr from the extract stream causes a problem [8]. ILs offer thedvantages of both homogenous and heterogeneous catalystsith their two main characteristics: A selected IL may be immis-
ible with the reactants and products, but on the other hand the ILay also dissolve the catalysts. ILs combine the advantages of a
olid for immobilizing the catalyst, and the advantages of a liq-id for allowing the catalyst to move freely [76]. Brennecke andaginn [8] indicated that the ionic nature of the IL also gives an
pportunity to control reaction chemistry, either by participatingn the reaction or stabilizing the highly polar or ionic transitiontates.
ILs have an active role in chemical reactions and catalysis.ome of the examples where ILs are utilized are: reactions ofromatic rings; clean polymerization [77]; Friedel Crafts alky-ation [78]; reduction of aromatic rings [79]; carbonylation [80];alogenation [81]; oxidation [82]; nitration [83]; sulfonation84]; solvents for transition metal catalysis; immobilization ofharged cationic transition metal catalysis in IL phase without
eed for special ligands [85]; in situ catalysis directly in ILather than aqueous catalysis followed by extraction of productsrom solution: this process eliminates washing steps, minimizesosses of catalysis and enhances purity of the products [86].
ritica
Mv
pigcrai
umartoehhrts
b[oodtItbdtfpa
5
oi[ciiiibtt[
ps
RtsftaBtieiicotp
nnipibIBitdb
ntbciovchibrfcravi
5
T
S. Keskin et al. / J. of Superc
any applications of ILs in catalytic reactions can be found inarious articles in the literature [1,12,85,87–89].
Holbrey and Seddon [90] described many of the catalyticrocesses which use low temperature ILs as reaction media andndicated that the classical transition metal catalyzed hydro-enation, hydroformylation, isomerization, dimerization andoupling reactions can be performed in IL solvents. In theireview, Holbrey and Seddon [90] concluded that ILs may be useds effective solvents and catalysts for clean chemical reactionsnstead of the volatile organic solvents.
Brennecke and Maginn [8], concluded that ILs have beensed successfully for hydrogenations, hydroformylations, iso-erizations, dimerizations, alkylations, Diels-Alder reactions
nd Heck and Suzuki coupling reactions, and in generalesearchers have concluded that the reaction rates and selec-ivities are as good or better in ILs than in conventionalrganic solvents. The catalytic hydrogenation of cyclohex-ne using rhodium-based homogenous catalysts [91] andydrogenation of olefins using ruthenium and cobalt-basedomogenous catalyst [92] in various ILs are studied and theesults indicated that there is a certain increase in the reac-ion rates and selectivity compared to the other normal liquidolvents.
Lagrost et al. [12] used immidazolium and ammonium-ased ILs ([emim][NTf2], [bmim][NTf2], [bmim][PF6],(C8H17)3NCH3][NTf2]) as reaction media for different typesf electrochemical reactions and investigated the oxidationf organic molecules (anthracene, naphthalene, durene, 1,4-ithiafulvene and veratrole) in ILs. Their results suggest thathe nature of investigated mechanisms is almost unchanged inLs as compared with the conventional organic media althoughhe structure of molecular solvents and ILs are expected toe quite different. Lagrost et al. [12] also concluded that theiffusion coefficient of the organic compounds are about 100imes smaller than those in conventional media as expectedrom the lower viscosity of RTILs versus organic solvents. Theositive results of this study demonstrated that ILs can be useds a new media for organic electrochemistry [12].
.4. Biological reactions media
ILs are used in biological reactions such as the synthesisf pharmaceuticals due to the stability of enzymes in ILs, andn separation processes such as the extraction of amino acids15]. IL biphasic systems are used to separate many biologi-ally important molecules such as carbohydrates, organic acidsncluding lactic acid [93,94]. Carbohydrates are renewable andnexpensive sources of energy and raw material for the chem-cal industry. The underivatized carbohydrates are not solublen most of the conventional solvents although they are solu-le in water. Their insolubility in most solvents prevents theransformation of carbohydrates. Therefore, the ability of ILso dissolve carbohydrates enables transformation possibilities
15,93,95,96].
Lau et al. [95] studied the alcoholysis, ammoniolysis, anderhydrolysis reactions by Candida antarctica lipase cataly-is using the [bmim[[PF6] and [bmim[[BF4] as reaction media.
rant
l Fluids 43 (2007) 150–180 159
eaction rates were generally comparable with, or better than,hose observed in organic media. Park and Kazlauskas et al. [96]tudied the acetylation of 1-phenylethanol catalyzed by lipaserom Pseudomonas cepacia (PCL) in several ILs and the reac-ion was as fast and as enantioselective in ILs as in toluene. Theylso investigated the acetylation of glucose catalyzed by lipase
from C. antarctica (CALB) and found that the transforma-ion was more regioselective in ionic liquids because glucoses up to one hundred times more soluble in ionic liquids. Liut al. [93] stated that carbohydrates are only sparingly solublen common organic solvents as well as in weakly coordinat-ng ionic liquids, such as [bmim][BF4]. They found that ILs thatontain the dicyanamide anion could dissolve approx. 200 g L−1
f glucose, sucrose, lactose and cyclodextrin and the esterifica-ion of sucrose with dodecanoic acid in [bmim][dca] could beerformed with CALB.
Swatloski et al. [97] showed that ILs can also be used ason-derivatizing solvents for cellulose, the most abundant biore-ewable material. Cellulose, which is insoluble in water andn most of the common organic solvents, has many derivitizedroducts in many applications of the fiber, paper, and polymerndustries. ILs incorporating anions which are strong hydrogenond acceptors are most effective solvents for cellulose, whereasLs containing non-coordinating anions including PF6
− andF4
− are not effective. Furthermore, Przybysz et al. [98] exam-ned the influence of ILs on a cellulose product, paper and foundhat the wettability of paper is improved, whereas the strengthecreased as a result of weakening of cellulose hydrogenonds.
Finally, Pfruender et al. [99] tested the water immiscible ILsamely; [bmim][PF6], [bmim][Tf2N] and [oma][Tf2N] (methyl-rioctylammonium bistrifluoromethanesulfonylimide) for theiriocompatibility towards Escherichia coli and Saccharomyceserevisiae. The results of this study showed that these watermmiscible ILs do not damage microbial cells and thereforene can utilize these water immiscible ILs as substrate reser-oirs and in situ product extracting agents for biphasic wholeell biocatalytic processes. Generally, toxic organic solventsave been used as substrate reservoirs and with this study its shown that water immiscible ILs may be used as biocompati-le solvents for microbial biotransformations. The experimentalesults demonstrated that there is an increase of chemical yieldrom <50% to 80–90% in simple batch processes and (R)-1-(4-hlorophenyl) ethanol was produced at a higher initial reactionate in the biphasic system (>50 �M s−1 L−1) compared to thequeous system [99]. Although ILs are known by their highlyiscous characteristics, good mass transfer rates were obtainedn their study.
.5. Removing of metal ions
Dai et al. [100] studied the effects of ILs (with PF6− and
f2N− anions) on improving the ability of crown ethers to
emove metal ions from aqueous solutions. Strontium nitrate,fission product for which there is no available extraction tech-ique for its removal from radioactive waste sites, was used inhis study.
1 ritica
rhtudd(Tnc[tati
ttavt
tolt[eo[eio
atbios
otbrdiadIs
maao
IToa
(o(Ktwur[hCT
6
tcwiop
wkocoablessllmIiitsd
mnIi
60 S. Keskin et al. / J. of Superc
Visser et al. [57] designed and synthesized several ILs toemove cadmium and mercury from contaminated water. Theydrophobic ILs come into contact with contaminated water andhey snatch the metal ions out of water. Task-specific ionic liq-ids (TSIL) concept is introduced in order to synthesize ILs withesired properties to extract metal ions. Visser et al. [57] pro-uced TSIL cations by appending different functional groupsnamely thiother, urea and thiourea) to imidazolium cations.hese IL cations can be considered as a new IL class, or aovel class of IL extractants. Synthesized TSIL cations wereombined with PF6
− anion and used alone or in a mixture withbmim][PF6]. The results of the study gave significant distribu-ion ratios for mercury and cadmium in liquid–liquid separationsnd minimized the reliance on traditional organic solvents forhis process. Davis [101] gave a detailed analysis and relatednformation on TSILs.
In traditional solvent extraction technologies, adding extrac-ants that reside quantitatively in the extracting phase increaseshe metal ion partitioning to the more hydrophobic phase. Thedded extractant molecules dehydrate the metal ions and pro-ide a more hydrophobic environment enabling their transporto the extracting phase [50].
In TSILs, attaching a metal ion coordinating group directlyo the imidazolium cation makes the extractant an integral partf the hydrophobic phase and in this way the chance for ILoss to the aqueous phase is reduced. Therefore, TSILs act ashe hydrophobic solvent and the extractant at the same time50]. However, the cost of TSIL is generally high. In order toliminate this drawback, TSILs may be added to the mixturesf less expensive ILs. Furthermore, Davis [101] and Zhao et al.15] stated in their recent reviews that TSILs are not limited toxtraction processes; they can be also used as versatile solventsn organic catalysts, solid phase synthesis and even in productionf liquid Teflon.
The extraction of radioactive metals (lanthanides andctinides) has particular industrial significance among IL extrac-ion of metal ions for the handling of nuclear materials [15]. Theehaviors of uranium species in various ILs were investigatedn early studies [102–105]. Recently, researchers have focusedn the fundamental understanding of ILs in nuclear chemistryuch as radiochemical stability of ILs [106].
Recently, Nakashima et al. [107] examined the feasibilityf extracting of rare earth metals into ILs from aqueous solu-ions and stripping of metal ions from ILs into an aqueous phasey complexing agents. They successfully accomplishing toecycle the extracting IL phase. In this study, octyl(phenyl)-N,N-iisobutylcarbamoylmethyl phosphine oxide (CMPO) dissolvedn [bmim][PF6] showed an extremely high extraction abilitynd selectivity of metal ions as compared to in an ordinaryiluent, n-dodecane. The results of this study indicate thatLs are a promising medium for actinide and fission producteparation.
In the literature, various studies were performed to extract
etal ions using ILs [108–113]. Different metal ions including
lkali, alkaline earth metals, heavy metals and radioactive metalsre researched by using different ILs. Generally, the side chainf the IL on the cation is varied and the effect of structure of the
3cN2
l Fluids 43 (2007) 150–180
L on the extraction efficiency of the metal ions is investigated.he side chain of the cation influences the hydrophobic characterf the IL and thus the partition coefficient of the metal ions isffected.
Visser et al. [108] extracted Na+, Cs+ using [Cnmim][PF6]n = 4, 6, 8); Chun et al. [109] investigated extraction ofther alkali metals such as Li+, K+, Rb+ using [Cnmim][PF6]n = 4–9); Luo et al. [112,113] studied the extraction of Na+,
+, Cs+ ions using [Cnmim][Tf2N] (n = 2, 4, 6, 8). Not onlyhe alkali metals but also extraction of alkaline earth metalsere studied by various groups: Visser et al. [108] removed Sr2+
sing [Cnmim][PF6] (n = 4,6,8); Bartsch et al. [110] studied theemoval of Mg2+, Ca2+, Sr2+, Ba2+; Luo et al. [112,113] utilizedCnmim][Tf2N] (n = 2, 4, 6, 8) to extract Sr2+. The extraction ofeavy and radioactive metals such as Cu2+, Ag+, Pb2+, Zn2+,d2+, Hg2+ were studied by using [Cnmim][PF6] (n = 4–9) andSILs [110,111,114].
. Challenges of ILs
The unique properties of ILs and the ability to designheir properties by choice of anion, cation and substituentsreate many more processing options, alternative to the onesith conventional solvents. However, high cost, lack of phys-
cal property and toxicity data restrict the advantageous usef ILs as process chemicals and processing aids at theresent.
The challenges in the use of ILs must be also addressed asell as their advantages. The major challenge is the cost. Ailogram of IL costed about 30,000-fold greater than a commonrganic solvent such as acetone. Renner [9] reported that this costould be reduced to approximately 1000-fold greater dependingn the composition of IL and the scale of production. Wagnernd Uerdingen [115] anticipated that the price of cation systemsased on imidazole will be in the range of D 50–100 kg−1, ifarger quantities of ILs are produced. The price can be loweredven below D 25 kg−1 if ILs are prepared with cheaper cationources on a ton scale. Another estimation was done by Wasser-heid and Haumann [116]. They expected that for ‘bulk ioniciquids” choosing proper (relatively cheap) cations and anionsead to prices approximately D 30 l−1 for production rates ofulti-ton. Moreover, scientists emphasized that the price of the
Ls may look still discouraging however, the essential factors the price to performance ratio. If the performance of an ILs extremely higher than that of the material (solvent) it aimso replace, less amounts of the IL may be needed for a givenpecific job [2], thus totally or partially overcoming the priceisadvantage.
The second problem is associated with the manufacturingethod of ILs. In manufacturing ILs environmental issues also
eed to be tackled since some VOCs are used to manufactureLs. Recently, some advanced methods have been developedn the solventless syntheses of ILs. For example, 1-alkyl-
-methylimidazolium halides have been synthesized in openontainers in a microwave oven without any VOCs by Varma andamboodiri at the Environmental Protection Agency of U.S.,001 [9].
ritica
tiptIosota
tfasitc
dacoga1fawe
ftiarwtlIbe
pputb
7
dotT
stedrs
idatSisbe
gfsfvIcffaIonalc
atcasnt
lffbmtbvant
S. Keskin et al. / J. of Superc
Researchers need to find alternative ways to recycle ILs due tohe reason that many processes for cleaning up ILs involve wash-ng with water or VOCs which creates another waste stream. Thisroblem has been solved by adopting supercritical extractionechnologies to recover the dissolved organic compounds fromLs or using membrane separation processes. However, there arether solid matrixes which adsorb some part of the ILs. Thus, aecond rinse would be required and this would create an aque-us waste stream that contains ILs. Despite this disadvantagehere may be some cleaning applications where ILs would bettractive [8].
Incomplete physico-chemical data are another challenge forhe application of ILs. At the present most available data areocused on bulk physical properties such as viscosity, densitynd phase transitions. Relatively little is known about the micro-copic physical properties of ILs. After these properties arenvestigated properly and the influence of ILs on chemical reac-ion rates is found, new ILs with precisely tailored propertiesan be synthesized.
It is extremely important to obtain reliable thermophysicalata and transport properties of ILs in order to make themvailable for many applications and to design IL-based pro-esses efficiently. Harris et al. [117] measured the viscositiesf two members of one of the most commonly studied ILroups, that are based on imidazolium cations; [omim][PF6]nd [omim][BF4] between 0 and 80 ◦C and at pressures to76 MPa ([omim][PF6]) and 224 MPa ([omim][BF4]) with aalling body viscometer and densities between 0 and 90 ◦Ct atmospheric pressure. The bulk physical properties of mostidely used ILs at wider temperature and pressure ranges are
ssential.Another barrier to the large-scale application of ILs arises
rom their high viscosities. The viscosities of ILs are higherhan most organic solvents and water, usually similar to viscos-ty of oils. This high viscosity may be responsible to produce
reduction in the rate of many organic reactions and even aeduction in the diffusion rates of species. Also, handling of ILsith high viscosities is difficult however; increasing tempera-
ure, changing anion–cation combinations may yield ILs withower viscosities. To overcome mass transfer limitations in gas-L systems resulting from high viscosity reactions using ILs maye run at high pressures and in efficient gas–liquid contactingquipment.
In chemical processing, pharmaceuticals, fine chemicals,etroleum refining, metal refining, polymer processing, pulp andaper, and textiles where a nonvolatile liquid with a wide liq-idus range could work better, ILs are the best choice however,he challenges of turning ILs into useful and environmentallyenign fluids must be overcome.
. ILs and scCO2 systems
Green chemistry, also known as sustainable chemistry,
escribes the search for reducing or even eliminating the usef substances in the production of chemical products and reac-ions which are hazardous to human health and environment.he goal of green chemistry is to create a cleaner and more
Ioit
l Fluids 43 (2007) 150–180 161
ustainable chemistry and it has received more and more atten-ion in recent years. Green chemistry searches for alternative,nvironmentally friendly reaction media as compared to the tra-itional organic solvents and at the same time aims at increasedeaction rates, lower reaction temperatures as well higherelectivities.
The ideal situation for a safe and green chemical processs using no solvent, however most of the chemical processesepend on solvents. Some of these solvents are soluble in waternd therefore they must be stripped from water before it leaveshe process not only for ecological but also for economic reasons.olvents must be recovered for recycle and reuse for an econom-
cally viable process. Water, perfluorinated hydrocarbons andupercritical fluids (SCFs) are alternative solvents which maye used in green chemistry. Among these, the most promisinglements of green chemistry are ILs and scCO2.
The low volatility of ILs is the key property that makes themreen solvents. However, this advantage also causes a problemor product separation and recovery [15]. Several techniques forolute recovery from ILs exist: volatile products can be extractedrom IL by distillation or simply by evaporation. However, non-olatile or thermo-sensitive products cannot be separated fromLs with these methods. ILs exhibiting immiscibility with wateran be extracted with water to separate water-soluble solutesrom IL into the aqueous phase; but this method is not suitableor hydrophilic ILs [17]. Of course, organic solvents such as hex-ne and toluene may be effective to recover the products fromL but this approach obviously compromises the ultimate goalf ‘green’ technologies [15]. Furthermore, the cross contami-ation between the phases presents another problem. Finally,nother green solvent is discovered which solves all the prob-ems and recovers various kind of solutes from ILs without crossontamination: supercritical fluids (SCFs).
SCFs are compounds which are above their critical temper-ture and pressure and they can be manipulated from gas likeo liquid like densities due to their unusual properties near theritical point. They are commercially viable solvents in severalpplications such as dry cleaning and polymer impregnation.cCO2 is the most widely used SCF as a result of nontoxic andon-flammable characteristics. scCO2 has low critical tempera-ure and pressure and it is not expensive.
The advantages of using SCFs as extraction medium includeow cost, nontoxic nature, recoverability and ease of separationrom the products. SCFs have been adapted for product recoveryrom ILs and supercritical fluid extraction (SCFE) is shown toe a viable technique with the additional benefits of environ-ental sustainability and pure product recovery [118]. Among
he SCFs, an inexpensive and readily available one, scCO2 hasecome a partner of IL and two environmentally benign sol-ents are utilized together in several applications. The volatilend nonpolar scCO2 forms different two-phase systems withonvolatile and polar ILs. The product recovery process withhese systems is based on the principle that scCO2 is soluble in
Ls, but ILs are not soluble in scCO2 [70]. Since most of therganic compounds are soluble in scCO2, with the high solubil-ty of scCO2 in ILs, these products are transferred from the ILo the supercritical phase.
162 S. Keskin et al. / J. of Supercritical Fluids 43 (2007) 150–180
Table 3IL–gas systems phase behaviors
System Temperature Pressure Findings Reference
[bmim][PF6]–CO2 40, 50, and 60 ◦C Up to 93 bar As pressure increases, solubility of CO2 in the IL increases [70][C8-mim][PF6]–CO2 Solubility of CO2 in IL-rich phase decreases with temperature[C8-mim][BF4]–CO2 CO2 solubility depends on the nature of the anion and cation[bmim][NO3]–CO2 The solubility of CO2 in IL-rich phase is highest for ILs with fluorinated anions[emim][EtSO4]–CO2 The general trend of the phase behavior is almost identical for all ILs[N-bupy][BF4]–CO2
[bmim][PF6]–CO2 10, 25, and 50 ◦C Up to 13 bar Water and carbon dioxide exhibited the strongest interactions and the [69][bmim][PF6]–C2H4 Highest solubilities in [bmim][PF6], followed by ethylene, ethane, and methane[bmim][PF6]–C2H6 Argon and oxygen both had very low solubilities and essentially no
interactions with the IL[bmim][PF6]–CH4
[bmim][PF6]–Ar[bmim][PF6]–O2
[bmim][PF6]–CO[bmim][PF6]–N2
[bmim][PF6]–H2
[bmim][PF6]–CO2 20, 40, 60, 80,100, and 120 ◦C
Up to 9.7 MPa Total pressure increases linearly with increasing amount of CO2 in IL [119]
[bmim][BF4]–CO2 30–70 ◦C Atmospheric pressure CO2 is found to be one order of magnitude more soluble in the IL than O2 [131][bmim][BF4]–O2 The solubility of CO2 in the IL decreases with temperature
The solubility of O2 in the IL slightly increases with temperatureSimilar results are obtained in [bmim][PF6]
[bmim][PF6]–CO2 25, 40, and 60 ◦C Up to 150 bar Solubility of CO2 in ten different IL is reported [121][bmim][BF4]–CO2 The solubility of CO2 is strongly dependent on the choice of the anion[bmim][TfO]–CO2 Increasing the alkyl chain length, increases the solubility of CO2 in IL[bmim][NO3]–CO2 All of the ILs expand a relatively small amount when CO2 is added[bmim][methide]–CO2
[bmim][DCA]–CO2
[bmim][Tf2N]–CO2
[hmim][Tf2N]–CO2
[omim][Tf2N]–CO2
[hmmim][Tf2N]–CO2
[emim][PF6]–CHF3 36.15–94.35 ◦C 1.6–51.6 MPa The solubility of supercritical CHF3 in [emim][PF6] is very high [142]At low CHF3 concentrations (mole fraction <0.5), the equilibrium pressureincreases almost linearly with CHF3 concentration, whereas further increase inCHF3 concentration causes a sharp increase in the equilibrium pressureThe Peng-Robinson EoS is capable of describing the experimental bubblepoint data of the system satisfactorily and qualitatively predicting the solubilityof the ionic liquid in supercritical CHF3
The high-pressure phase behavior of [emim][PF6]–CHF3 system wascompletely different from that of [bmim][PF6]–CO2
[emim][PF6]–CO2 35–93 ◦C 1.49–97.10 MPa CO2 is more soluble in [bmim][PF6] than in [emim][PF6] [129]The general phase behavior of [emim][PF6]–CO2 and [bmim][PF6]–CO2
systems are found to be completely similarCO2 is more soluble in [bmim][PF6] than in [emim][PF6]The phase behaviors of the systems [emim][PF6]–CO2 and[emim][PF6]–CHF3 are differentCHF3 is more soluble in the IL than CO2 at higher pressures and the IL is alsomore soluble in supercritical CHF3 than in supercritical CO2
[hmim][PF6]–CO2 25.16–90.43 ◦C 0.64–94.60 MPa CO2 is more soluble in [hmim][PF6] than in [emim][PF6] [130]The general phase behaviors of [emim][PF6]–CO2 and [hmim][PF6]–CO2 aresimilarThe solubility of the IL in scCO2 phase is very low and cannot be detected
[bmim][PF6]–CO2 25 ◦C Up to 1 MPa A group contribution form of a non-random lattice fluid model is applied topredict the solubility of CO2 in ILs
[123][C6mim][PF6]–CO2
[emim][BF4]–CO2
[C6mim][BF4]–CO2
[emim][Tf2N]–CO2
[C6mim][Tf2N]–CO2
S. Keskin et al. / J. of Supercritical Fluids 43 (2007) 150–180 163
Table 3 (Continued )
System Temperature Pressure Findings Reference
[bmim][PF6]–CO2 40–90 ◦C Up to 97 MPa CO2 has good solubilities in these ILs at lower pressures [13][emim][PF6]–CO2 There is a linear relationship between the alkyl chain length and solubility
of CO2[hmim][PF6]–CO2
[hmim][BF4]–CO2 20–95 ◦C 0.54–100 MPa CO2 was found to be more soluble in [hmim][PF6] than in [hmim][BF4] [133][hmim][PF6]–CO2
[bmim][BF4]–CO2 5.32–95.07 ◦C 0.587–67.62 MPa CO2 has a high solubility in [bmim][BF4] at lower pressures, but the solubilitydecreases dramatically at higher pressures
[132]
The phase behavior of the system [bmim][BF4]–CO2 shows similarities withthe phase behavior of the system [hmim][BF4]–CO2
CO2 is more soluble in [hmim][BF4] than in [bmim][BF4]
[omim][BF4]–CO2 29.85–89.95 ◦C 0.1–100 MPa High solubilities of CO2 for low CO2 mole fractions (mole fraction < 0.6) areattained at relatively low pressure, whereas for mole fraction >0.6, the pressureneeded for completely dissolving the CO2 increases drastically
[139]
Increase in temperature slightly decreases the solubility of CO2
The CO2 solubility increases in the IL with increasing chain length of the alkylgroup
for fof CO
7
vkantp
saar[eptpo
fia
7
ecmis
vbp
tgksWoCCihtaDAaiYowr
5id0iirtiio
Suggestedsolubility
.1. High-pressure phase behavior of IL–CO2 systems
Preliminary works have shown that scCO2 extraction is aiable method for solute recovery from an IL. However, thenowledge of phase behavior of IL–CO2 systems is an essentialspect of this methodology. scCO2 dissolution in the IL phase isot only necessary for contact with the solute but it also reduceshe viscosity of the IL and therefore enhancing the mass transferrocess.
Early studies of IL–CO2 phase behavior indicated that theseystems are very unusual biphasic systems. No measurablemount of [bmim][PF6] was soluble in the CO2-rich phase,lthough a large amount of CO2 dissolved in the IL-rich phase,educing the viscosity of IL [70]. Blanchard and Brennecke118] concluded that the system remained as two distinct phasesven under pressures up to 400 bar. Therefore, high-pressurehase behavior of [bmim][PF6]–CO2 is totally different fromhat of any ordinary organic liquid–CO2 systems. This differenthase behavior is the key phenomena which makes extractionf solutes from IL with CO2 attractive.
Table 3 summarizes the phase behavior studies performedor IL–gas systems, demonstrates the type of IL and gases usedn these studies, the experimental conditions, the basic findingsnd the related references.
.1.1. The [bmim][PF6]–CO2 systemThe phase behaviors of IL–scCO2 systems are studied very
xtensively in the literature for a better understanding of the pro-esses involving both IL and scCO2. Since [bmim][PF6] is theost widely studied IL in the literature, many researchers stud-
ed the high-pressure phase behavior of the [bmim][PF6]–CO2ystem [13,69,70,119–122].
Blanchard et al. [70] measured the high-pressureapor–liquid phase behavior of [bmim][PF6]–CO2 systemy using two different apparatus sets: a static high-pressurehase equilibrium apparatus and a dynamic flow apparatus. In
Cbpi
urther study to investigate the effect of the anion on the
2.when the cation is fixed as [omim]
he static high-pressure vapor–liquid equilibrium apparatus, alass cell was loaded with a known amount of IL sample andnown amounts of CO2 were metered into the cell while theample within was vigorously stirred to ensure equilibrium.
ith the assumption of pure CO2 vapor phase, the compositionf the IL-rich phase was calculated by knowing the amount ofO2 added to the cell. At the end of the equilibration period,O2 was completely removed from IL phase upon depressur-
zation. The same group also used a dynamic apparatus, i.e. aigh-pressure extractor to determine the solubility of the IL inhe CO2 phase. The detailed description of these apparatus setsnd experimental procedures can be found in the literature [70].ifferent experimental set-ups were used by other researchers:
schematic diagram of a general IL–scCO2 experimentalpparatus, which was used to measure the solubility of CO2n IL ([bmim][PF6]) is given by Kim et al. [123]. Shiflett andokozeki [124] measured the gas solubility and diffusivityf CO2 in [bmim][PF6] using a gravimetric microbalance forhich the details of the experimental set-up is given in the
elated reference.The solubility of CO2 in [bmim][PF6] was determined at 40,
0 and 60 ◦C and pressures up to 93 bar [70]. As the pressurencreases, the solubility of CO2 in the IL-rich phase increasesramatically and the solubility value reaches a mole fraction of.72 at 40 ◦C and 93 bar. A general rule suggests that an increasen temperature results with a decrease in the solubility of gasesn liquids. As expected, the solubility of CO2 in [bmim][PF6]ich phase decreases with temperature. However, they noticedhat the temperature dependence of the solubility is quite smalln this temperature and pressure range. Another crucial points the effect of large degree of CO2 solubility on the viscosityf IL. The viscosity of IL decreases when a certain amount of
O2 is dissolved in IL and this effect can easily be observedy the reduced drag on the stirring magnet when a static high-ressure vapor–liquid equilibrium set-up is used. This reductionn viscosity of the liquid facilitates the solution process.
164 S. Keskin et al. / J. of Supercritical Fluids 43 (2007) 150–180
F(S
[pspldro
vtioaBit[i
Faw
FpC
wITi[d
bpFfoEv
ig. 7. Total pressure versus molality of gas for [bmim][PF6]–CO2 system.Reprinted with permission from [119]. Copyright 2003 American Chemicalociety)
Kamps et al. [119] presented the solubility of CO2 inbmim][PF6] for temperatures 293–393 K in 20 K intervals andressures up to about 9.7 MPa. The total pressure is plotted ver-us the stoichiometric molality of the gas (number of moleser kilogram of the IL) in Fig. 7. The total pressure increasesinearly with increasing amount of the gas in IL. The solubilityata represented by Kamps et al. [119] differ from the previouslyeported solubility data of Blanchard et al. [70]. The comparisonf experimental data of two studies is given in Fig. 8.
There are several [bmim][PF6]–CO2 high-pressureapor–liquid equilibrium data sets available in the litera-ure. However, these sets differ from each other considerablyn the values they report for similar conditions. The reasonf differing solubility data reported may be due to the smallmounts of water dissolved in the IL sample used. For example,lanchard et al. [125] presented a solubility data of CO2
n [bmim][PF6] which is different than the data reported byhe same group in 2001. In the first study, this group usedbmim][PF6] which was saturated with water at 22 ◦C, contain-ng 2.3 wt.% water. In the second study, they used [bmim][PF6]
ig. 9. [bmim][PF6]–CO2 liquid phase compositions for dried and wet IL sam-les at 40 ◦C. (Reprinted with permission from [70]. Copyright 2001 Americanhemical Society)
hich was dried to approximately 0.15 wt.% water. Drying ofL has a significant effect on the phase behavior with CO2.hus, Blanchard et al. [70] showed that the solubility of CO2
n ILs was decreased in the presence of water. In Fig. 9,bmim][PF6]–CO2 liquid phase compositions are given forried and wet IL samples.
In order to observe the effect of water impurity in IL, phaseehaviors of two IL samples (dry and wet) with CO2 were com-ared. The effect of water impurity in IL is significant at 57 bar.or dried IL sample, the mole fraction of CO2 is 0.54, whereasor the wet (water saturated) IL sample it is only 0.13. The effectf water in IL may be explained by CO2-phobic nature of water.ven at high pressures, mutual solubilities of water and CO2 areery low [126]. Another point is the formation of carbonic acidrom the reaction of CO2 with water that can result in a reductionf the aqueous phase pH to as little as 2.80 [127].
Rubero and Baldelli [128] investigated gas–liquid interfacef imidazolium ILs using surface-sensitive vibrational spec-roscopy sum frequency generation. The results indicated thathen the IL is dry, the cation is oriented with the imida-
olium ring parallel to the surface plane for both hydrophilicnd hydrophobic ILs. But the cation reorients itself with respecto the surface for the hydrophobic liquid when water is added,hile the orientation in the hydrophilic liquid is unaffected.After the influence of water is noticed, researchers working
n IL–CO2 solubility and equilibrium have started to dry andegas all ILs under vacuum at room temperature for severalays prior to use. After ILs are dried and degassed, the waterontents are estimated by the Karl Fischer analysis before solu-ility data are taken. Measurements show that the most widelytudied IL; [bmim][PF6] absorbs a couple wt.% water when lefto the atmosphere. The estimated water content of [bmim][PF6]
fter drying was approximately 0.15 wt.% water as measured byarl Fischer analysis [70].A number of high and low-pressure [bmim][PF6]–CO2 solu-
ility studies have appeared in the literature. Although consistent
ritica
rCpTpu
oI[ptwb
[nerahaasbCtcrCp
cmgsbtt[aa
Cid[r[a[oagstp
fat
mat[ht(Kst[i
t[m[ftaacl[uafd[[
toipeoBSowaobaBaho
S. Keskin et al. / J. of Superc
esults have been established for low-pressure solubility data ofO2 in [bmim][PF6], there are large discrepancies among high-ressure solubility data of several researchers [13,70,119–121].hese large solubility differences in the literature are mostrobably due to the differences in the purities of the ILssed.
Since the behaviors of IL–CO2 systems are different fromther organic liquid–CO2 systems, full phase diagrams ofL–CO2 systems are investigated. Blanchard and co-workers125] found two-phase immiscibility regions with three cloudoint measurements of 1.31, 4.92 and 7.15 mole% IL mix-ures with the balance being CO2. Although, these experimentsere conducted with water-saturated ILs, qualitatively a similarehavior is expected with dried samples.
Blanchard et al. [70] gave a qualitative phase behavior ofbmim][PF6]–CO2 system over a wide pressure range. Theyoticed that the phase behavior where a large miscibility gapxists even at extremely high pressures. Blanchard et al. [70]eported and referred that as a complementary work, McHughnd co-workers studied [bmim][PF6]–CO2 phase behavior atigher pressures up to 3100 bar and found two distinct phasest all conditions. The existence of large immiscibility gap event very high pressures is not expected for organic liquid–CO2ystems and the existence of two distinct phases is explainedy the following discussion: At high-pressures density of pureO2 phase increases but since the liquid phase does not expand,
he two phases will never become identical and a mixture criti-al point will never be reached. Therefore, the IL–CO2 systememains as two phases even at very high pressures, although theO2 solubility is quite high, the mixtures never become a singlehase [70].
Anthony et al. [69] reported the solubilities and Henry’sonstants of different gases (carbon dioxide, ethylene, ethane,ethane, argon, oxygen, carbon monoxide, hydrogen, and nitro-
en) in [bmim][PF6] and showed that CO2 has the highestolubility and strong interaction with [bmim][PF6]. The solu-ility data of CO2 in [bmim][PF6] is in good agreement withhe published results of Blanchard et al. [70] although differentechniques were used in these studies. Furthermore, Baltus et al.68] reported that Henry’s constants for Kamps et al. [119] datare in reasonable agreement with those obtained by Anthony etl. [69].
Aki et al. [121] studied the high-pressure phase behavior ofO2 in imidazolium-based ILs and compared the phase behav-
or of the system [bmim][PF6]–CO2 with the other solubilityata present in the literature. The solubility results of CO2 inbmim][PF6] at 25 ◦C measured by Aki et al. [121] agreedemarkably well with the solubility results of Anthony et al.69] at low pressures and with the solubility results of Kamps etl. [119]. Aki et al. [121] investigated the solubility of CO2 inbmim][PF6] at 40 ◦C and compared the results with the previ-us studies of Kamps et al. [119], Blanchard et al. [70] and Liu etl. [120]. As expected the agreement between the data points is
ood at low pressures but the discrepancy is obvious at high pres-ures. Aki et al. [121] explained that in their previous work [70],hey were not aware of the various impurities and degradationroducts that were present in the samples they used. There-
tCro
l Fluids 43 (2007) 150–180 165
ore, they attributed the difference between their study [121]nd the previous study of the same group [70] to the purity ofhe IL.
At 40 ◦C and at all pressures, there is a very good agree-ent within the solubility values reported by Aki et al. [121]
nd Liu et al. [120]. However, this statement is not correct forhe reported solubility data of Aki et al. [121] and Kamps et al.119]. The results of two studies agree at low pressures, but atigh pressures, there is a significant difference: At about 43 bar,he solubility of CO2 in [bmim][PF6] was measured as 0.43mole fraction) by Aki et al. [121], however, at the same pointamps et al. [119] reported the solubility of CO2 as 0.38. By con-
idering the studies mentioned above, it may thus be concludedhat the solubility of CO2 in one of the most widely studied IL,bmim][PF6], varies among different groups in the literature ands not a good agreement especially at higher pressures.
The solubility of CO2 in [bmim][PF6] was experimen-ally studied at 298.15 K and up to 1.0 MPa by Kim et al.123]. A group contribution form of a non-random lattice–fluidodel (GC-NLF) was applied to predict solubility of CO2 in
bmim][PF6]. They used the solubility data of Kamps et al. [119]or a wider pressure range for the group parameter determina-ion. Comparisons of calculated solubility data with Kamps etl. data [119] for [bmim][PF6] demonstrated that the methodpplied is fairly accurate except for regions close to the criticalonditions of CO2. Kim et al. [123] also compared the calcu-ated values of solubility of CO2 in different ILs ([emim][PF6],bmim][PF6], and [C6mim][PF6]) with the experimental sol-bility data reported by other groups for the same ILs. Thegreements are generally good up to 10 MPa pressure, however,urther comparisons for higher pressure shows some degree ofiscrepancy between the calculated solubility data of Kim et al.123] and the experimental solubility data of Shariati and Peters129,130].
Finally, Shariati and Peters [13] studied the comparison ofhe phase behavior of [bmim][PF6]–scCO2 system with thether studies present in the literature. The solubility of CO2n [bmim][PF6] was determined by measuring the bubble pointressure of the binary system at different temperatures for sev-ral isopleths and pressures up to 97 MPa. The solubility dataf CO2 in [bmim][PF6] at 323.15 K was compared with that oflanchard et al. [70] and Anthony et al. [71] and the results ofhariati and Peters [13] were in a good agreement with thosef Anthony et al. [71]. The solubility data taken at 333.15 Kas compared with that of Blanchard et al. [70], Kamps et
l. [119], Liu et al. [120]. Although the experimental meth-ds were completely different, there is also a good agreementetween the results of Shariati and Peters [13] and Kamps etl. [119] at a temperature of 333.15 K. The solubility data oflanchard et al. [70] and Liu et al. [120] show greater devi-tions from the data of Shariati and Peters [13] especially atigher pressures. Shariati and Peters [13] reported the existencef the three-phase equilibrium liquid–liquid–vapor (LLV). As
he other studies demonstrated, this study also indicated thatO2 has a high solubility in [bmim][PF6] and there is a linear
elationship between the alkyl chain length and the solubilityf CO2.
1 ritica
7
IaChd3n3Ni5wbugobia[
svf0Atnp[osb
bc[t8s[mIsil
wroprsaK
Iissatsecb(gatii
biritbSes[wi[
tauia[nrfttdeKit
aIaCr
66 S. Keskin et al. / J. of Superc
.1.2. Other IL–CO2 systemsHigh-pressure phase behavior of different types of
Ls are similar to that of [bmim][PF6]. Blanchard etl. [70] investigated the high-pressure phase behavior ofO2 with six different ILs: 1-n-butyl-3-methylimidazoliumexafluorophosphate ([bmim][PF6]), 1-n-octyl-3 methylimi-azolium hexafluorophosphate ([C8-mim][PF6]), 1-n-octyl--methylimidazolium tetrafluoroborate ([C8-mim][BF4]), 1--butyl-3-methylimidazolium nitrate ([bmim][NO3]), 1-ethyl--methylimidazolium ethyl sulfate ([emim][EtSO4]), and-butylpyridinium tetrafluoroborate ([N-bupy][BF4]). They
nvestigated the solubility of CO2 in different ILs at 40,0, 60 ◦C and pressures up to 93 bar. The focus of theork was to develop an insight into the physical interactionetween CO2 and ILs with different cation–anion config-rations. The solubility of CO2 in the IL-rich phase wasreatest for ILs with fluorinated anions, following the trendf [bmim][PF6] and [C8-mim][PF6] > [C8-mim][BF4] > [N-upy][BF4] > [bmim][NO3] > [emim][EtSO4] [70]. The solubil-ty of CO2 was found to be greatest in ILs with PF6
−, anionnd the IL that exhibited the lowest solubility of CO2 wasemim][EtSO4].
The solubility of CO2 in ILs increases with increasing pres-ure but the exact amount of CO2 dissolved in the liquid phasearies significantly. In the study of Blanchard et al. [70], it isound that at 70 bar the solubility of CO2 in [emim][EtSO4] was.36 (mole fraction) whereas, it was 0.63 in [C8-mim][PF6].lthough there are some numerical differences in the mole frac-
ion of CO2 dissolved, the general trend of the phase behavior isearly identical for all ILs. The solubility of CO2 in the IL-richhase changes slightly with the temperature as in the case ofbmim][PF6]–CO2 system. Since the qualitative phase behaviorf almost all ILs seems similar, it can easily be concluded thatcCO2 can be used to recover solutes not only from [bmim][PF6]ut also from all kind of ILs.
In order to understand the effect of the anion on the phaseehavior of IL–CO2 systems, two pairs of ILs with the sameations were compared: [C8mim][PF6]–[C8mim][BF4] andbmim][PF6]–[bmim][NO3]. Changing the anion from [PF6]−o [BF4]− in the [C8mim] salts, results in an approximately% decrease in CO2 solubility at 40 ◦C over the range of pres-ures studied by Blanchard et al. [70]. The solubility of CO2 inbmim][NO3] is about 25% less than in [bmim][PF6]. Experi-ental and molecular simulation studies found that the anions of
L dominate the interaction with CO2, with the cation playing aecondary role. Cadena et al. [122] pointed out that the changesn the imidazolium cation involving alkyl groups have relativelyittle influence on the solubility of CO2 in IL.
[bmim][BF4] is one of the popular ILs which has been veryidely used in most of the studies: Husson-Borg et al. [131]
eported the solubility of CO2 in [bmim][BF4] as a functionf temperature between 303 and 343 K and at atmosphericressure. This group used a new type of experimental appa-
atus based on a saturation method. The equilibrium cell ispecially designed for viscous solvents like the IL, and anppropriate gas–liquid contact is obtained by good agitation.roon et al. [132] studied the phase behavior of different
sC
[
l Fluids 43 (2007) 150–180
L-supercritical fluid systems including [bmim][BF4]. Theynvestigated the phase behavior of [bmim][PF6]–CO2 binaryystem experimentally and reported its bubble point pres-ures for CO2 concentrations between 10.22 and 60.17 mole%nd in a temperatures range of 278.47–368.22 K. They foundhat CO2 has a high solubility at lower pressures, but theolubility decreases dramatically at higher pressures. Thexperimental results for the [bmim][BF4]–CO2 system wereompared with the available phase behavior data of theinary system 1-hexyl-3-methylimidazolium tetrafluoroborate[hmim][BF4])–CO2 [133] to investigate the effect of the alkylroup length on the phase behavior. The results showed thatlarger alkyl group led to lower bubble-point pressures and,
herefore, to higher solubilities of CO2 in the imidazolium-basedonic liquid. Thus, CO2 was more soluble in [hmim][BF4] thann [bmim][BF4].
Shariati and Peters [129] studied the phase behavior ofinary system [emim][PF6]–CO2 experimentally by measur-ng its bubble point pressures at temperatures and pressureanges of 308.14–366.03 K and 1.49–97.10 MPa, and comparedt with [bmim][PF6]–CO2 system to understand the effect ofhe length of alkyl chain on the solubility of CO2. The solu-ility of CO2 in [bmim][PF6] is higher than in [emim][PF6].ince the butyl group in [bmim][PF6] is bulkier than thethyl group in [emim][PF6], Shariati and Peters [129] empha-ized that CO2 can dissolve better in [bmim][PF6] than inemim][PF6], the latter component being denser. Moreover, itas observed that the experimentally determined phase behav-
or of the [emim][PF6]–CO2 is similar to the phase behavior ofbmim][PF6]–CO2 at 333.15 K.
Aki et al. [121] also presented the solubility of CO2 inen different imidazolium-based ILs at 25, 40, and 60 ◦Cnd pressures to 150 bar. They concluded that the sol-bility of CO2 in imidazolium-based ILs increases withncreasing pressure and decreases with increasing temper-ture for all the ILs investigated. Furthermore, Aki et al.121] investigated the influence of the different anions,amely; dicyanamide ([DCA]), nitrate ([NO3]), tetrafluorobo-ate ([BF4]), hexafluorophosphate ([PF6]), trifuoromethanesul-onate ([TfO]), bis(trifluoromethylsulfonyl) imide ([Tf2N]), andris(trifluoromethylsulfonyl)methide ([methide]). The results ofheir study showed that the solubility of CO2 is strongly depen-ent on the choice of anion. Solubility measurements of Anthonyt al. [69] and Cadena et al. [122]; spectroscopic studies ofazarian et al. [134] also all showed that the solubility of CO2
n ILs depend on the anion, especially the interaction betweenhe anion of the IL and CO2.
CO2 is the least soluble in the two ILs with non-fluorinatednions, [NO3] and [DCA], and it has the highest solubility inLs with anions containing fluoroalkyl groups, [TfO], [Tf2N],nd [methide]. Aki et al. [121] attributed the high solubility ofO2 in ILs to the anions containing fluoroalkyl groups as a
esult of favorable interactions between CO2 and the fluoroalkyl
ubstituents on the anion since fluoroalkyl groups are known asO2-philic ones [135–137].
Aki et al. [121] also compared the solubility of CO2 inbmim][Tf2N], [hmim][Tf2N], and [omim][Tf2N] at 25, 40, and
ritica
6TcacPtwIdivIceit
art3spi[gt
trmat[ClfodipE3cop
owleptltd
ssso
tnP[t
7
IicCdCa[[seTto
amwdiInmts[sssuomrwthsot
S. Keskin et al. / J. of Superc
0 ◦C to investigate the influence of cation alkyl chain length.he solubility of CO2 increases with an increase in the alkylhain length at all pressures, with the increase being morepparent at higher pressures. This result of Aki et al. [121] isonsistent with the results of previous studies of Shariati andeters [129,130]. Furthermore, Blanchard et al. [70] reported
hat the CO2 solubility increases when the alkyl chain lengthas increased from butyl to octyl for ILs containing [PF6].
t is known that the densities of the imidazolium-based ILsecrease as the alkyl chain length increases [49,138]. Therefore,midazolium-based ILs with longer alkyl chains have greater freeolume and thus greater solubility of CO2 is expected in theseLs. Therefore, the higher solubility of CO2 in ILs with longeration alkyl chains may be attributed to the entropic rather thannthalpic arguments [121]. Finally, it may be concluded that its possible to increase the solubility of CO2 in ILs by increasinghe alkyl chain length on the cation.
Constantini et al. [133] investigated the phase behavior ofbinary mixture of 1-hexyl-3-methylimidazolium tetrafluo-
oborate ([hmim][BF4])–scCO2 system and compared it withhe experimental data of the binary system of 1-hexyl--methylimidazolium hexafluoroborate ([hmim][PF6])–scCO2ystem, in order to demonstrate the anion effect. Although thehase behaviors of the binary systems are similar, the solubil-ty of the scCO2 is found to be higher in [hmim][PF6] than inhmim][BF4]. This higher solubility may be explained with thereater interaction between CO2 and the [PF6] anion, althoughhe former is denser than the latter [133].
In another study [139], the high-pressure phase behavior ofhe binary system 1-octyl-3-methylimidazolium tetrafluorobo-ate ([omim][BF4])–CO2 was studied in the liquid-phase CO2ole fraction range 0.1–0.75, and in the pressure and temper-
ture range of 0.1–100 MPa and 303–363 K. They comparedheir experimental data with the data given by Blanchard et al.70]. Although there was a discrepancy between two studies forO2 mole fractions higher than 0.6, they both suggested a pecu-
iar behavior of a rapid pressure increase at higher CO2 moleractions [139]. Kroon et al. [140] presented the phase behaviorf several IL–CO2 binary systems experimentally, and finallyeveloped an equation of state (EoS) to predict the phase behav-or of IL–CO2 systems based on the truncated perturbed chainolar statistical associating fluid theory (tPC-PSAFT) EoS. TheoS was used to describe the CO2 solubility in several 1-alkyl--methylimidazolium-based ionic liquids with different alkylhain lengths within a pressure and CO2 mole fraction rangef 0–100 MPa and 0–75%, respectively. The binary interactionarameter was fitted to V–L equilibrium data.
In general when a gas is dissolved in a liquid phase, dilationf the liquid occurs. ILs show only slight dilations in volumeith CO2 dissolution. This behavior is different than normal
iquid–CO2 systems. In fact, the dissolution of CO2 in liquids toxpand them and reduce their solvent strength is a well-knownhenomena and it is the basis of gas anti-solvent (GAS) process
o precipitate solutes from liquids [141]. However, ILs do not fol-ow this trend due to the strong Coulombic forces associated withhe ionic nature of ILs. The lack of expansion of ILs seems toerive from the fact that dissolved CO2 does not greatly affect the
oats
l Fluids 43 (2007) 150–180 167
trength of interionic interactions in IL. Therefore, the IL–CO2ystem is extremely different from other organic liquid–CO2ystems with the large immiscibility region and lack of dilationf the liquid phase [70].
Although there are not so many available studies concerninghe IL/‘SCF other than CO2
′ systems, for the complete-ess of the material they must be mentioned. Shariati andeters [142] investigated the high-pressure phase behavior ofemim][PF6]–supercritical fluoroform, the experimental condi-ions and the results of this study is given on Table 3.
.2. IL solubility in CO2
The majority of the research focus on the solubility of CO2 inL, on the other hand the concentration of IL in CO2-rich phases also important. A dynamic apparatus with a high-pressureell, was used to measure the solubility of [bmim][PF6] in theO2-rich phase [70]. The solubility of [bmim][PF6] in CO2 wasetermined at 40 ◦C and 137.9 bar by flowing 0.5866 mole ofO2 through a cartridge loaded with [bmim][PF6]. UV–visiblenalysis gave no appreciable IL absorption peak, indicatingbmim][PF6] solubility of less than 5 × 10−7 (mole fraction ofbmim][PF6]) of in the CO2 phase. Due to the fact that no mea-urable IL dissolves in CO2, a solute dissolved in an IL can beasily recovered with scCO2 without any cross contamination.he lack of solubility of IL in the CO2 phase can be attributed to
wo reasons: extremely low vapor pressure of IL and the inabilityf CO2 to adequately solvate ions in the gaseous phase.
Although the solubility of IL in scCO2 is extremely lownd not measurable, in industrial applications, the scCO2 phaseay contain some other components such as reactants, productshich may act as cosolvents to enhance the ability of scCO2 toissolve IL significantly. In this case, the amount of IL dissolvedn CO2-rich phase may not be negligible under some conditions.n order to decide on the conditions to avoid cross contami-ation, solubility of IL in scCO2/organic compound mixturesust be known. Wu et al. [143] conducted the first study in
he literature to observe the effect of organic compounds incCO2 on the solubility of an IL in CO2 phase. The solubility ofbmim][PF6] in scCO2, and in scCO2–ethanol, scCO2–acetone,cCO2–n-hexane mixtures was investigated quantitatively. Theolubility of IL in scCO2 is extremely low as the earlier studiesuggested, however by addition of ethanol and acetone, the sol-bility of IL increases dramatically as the concentration of therganic compounds in scCO2 exceeds 10 mole%. This enhance-ent of the IL solubility by addition of ethanol and acetone
esults mainly from strong interaction of the two compoundsith the IL due to their strong polarity. The polarity of ace-
one is stronger than ethanol; therefore, the solubility of IL isigher in the scCO2–acetone system than in the scCO2–ethanolystem. Since n-hexane is a nonpolar substance its influencen the solubility of IL in scCO2 phase is very limited. A fur-her study of Wu et al. [144] also showed that the ability
f cosolvents to increase the solubility of ILs ([bmim][PF6]nd [bmim][BF4]) in scCO2 follows the order: acetoni-rile > acetone > methanol > ethanol > n-hexane. This order isame with the order of dipole moments of cosolvents. With this
168 S. Keskin et al. / J. of Supercritical Fluids 43 (2007) 150–180
Table 4IL solubility in CO2
System Temperature Pressure Findings Reference
[bmim][PF6]–CO2 40 ◦C 137.9 bar The solubility of [bmim][PF6] in scCO2 is extremely low [70]scCO2 may be used to recover solutes dissolved in ILwithout any cross contamination
[bmim][PF6]–CO2 40 and 55 ◦C 12–15 MPa By addition of ethanol and acetone, the solubility of the ILin scCO2 phase enhances
iscussion, it is clear that the polarity of the organic compound isdominant factor in influencing solubility. With this study Wu etl. [144] emphasized that the amount of IL dissolved in scCO2-ich phase may be significant if the system contains sufficientlyolar organic compounds in sufficient concentrations. Table 4hows the studies of Blanchard et al. [70], Wu et al. [143] and
u et al. [144], the components of their systems, experimentalonditions and basic results.
The phase behavior of the IL–CO2–methanol system andhe viscosity of the mixtures were studied previously byiu et al. [120] at different conditions. The phase behaviorf IL–CO2–water system was investigated by Zhang et al.145] and finally this group studied the phase behavior ofbmim][PF6]–CO2–acetone system in detail at 313.15 K over aide pressure range. Zhang et al. [146] determined the distribu-
ion coefficients of the components between different phases andound that CO2 distribution coefficient decreases with increas-ng pressure while the acetone distribution coefficient increasedith pressure.
.3. IL–CO2 interaction at the molecular level
An in situ attenuated total reflectance–infra red (ATR–IR)tudy of CO2 dissolved in two ILs ([bmim][PF6] andbmim][BF4]) at high pressures has demonstrated the effectsf anionic species of the ILs on the molecular state of the dis-olved CO2. Kazarian et al. [147] showed that CO2 forms weakewis acid–base complexes with the anions in [bmim][PF6] and
bmim][BF4]. Furthermore, they demonstrated that this interac-ion is stronger with [bmim][BF4]. BF4
− acts as a stronger Lewisase towards CO2 than PF6
−. In addition to this, as the size ofnion increases the strength of interaction decreases. However,he solubility data of various group indicated that CO2 has aigher solubility in [bmim][PF6] than in [bmim][BF4]. Thus,
he strength of these interactions cannot be solely responsibleor the solubility of CO2 in these ILs, and presumably, a freeolume contribution in IL plays a significant role [148]. Thetrength of the anion–cation interactions in IL affects the avail-
asco
ble free volume and one can expect that a weaker interactingnion leads to more free volume being available.
ILs are generally distinguished with their low melting points.nother in situ ATR–IR spectroscopic study of Kazarian et al.
134] showed that high-pressure CO2 reduces the melting tem-erature of ILs. This possibility of reducing melting temperaturef ILs further under high pressure CO2 provides a new oppor-unity to use ILs as solvents at milder temperatures. Kazarian etl. [134] investigated the effect of CO2 pressure on the phaseehavior of 1-hexadecyl-3-imidazolium hexafluorophosphate.he data presented indicated that 70 bar CO2 reduces the melt-
ng point of this IL from 75 to 50 ◦C. This result was assignedo a weak Lewis acid–base type interaction between anion andO2, with the P–F bonds perpendicular to O C C axis thereby
educing the rather stronger interactions between the P–F bondsnd the cations. CO2 disrupts the cation–anion and the tail-tailnteractions in IL rather than simply playing an impurity role inhe mechanism of induced melting. Preliminary results obtainedy Kazarian et al. [134] also indicated that melting temperaturesf other analogous ILs decrease by high pressure CO2 allowingheir use in many applications at mild temperatures.
.4. Solute recovery from ILs with scCO2
Researchers found that nonvolatile organic compounds cane extracted from ILs using scCO2, which is widely used toxtract large organic compounds with minimal pollution. Blan-hard et al. [125] showed that CO2 can be used to extractaphthalene, a low volatility model solute, from an IL. Theyynthesized [bmim][PF6] which was stable in the presence ofater and oxygen. The model compound naphthalene was read-
ly soluble in [bmim][PF6] (maximum solubility of 0.30 moleraction at 40 ◦C) and in CO2. Their study investigated the phaseehavior of [bmim][PF6] with CO2, as well as with naphthalene
nd finally that of the [bmim][PF6]–CO2–naphthalene ternaryystem. The results showed that CO2-rich phase was not signifi-antly contaminated by IL, as would be expected during contactf CO2 with any conventional organic solvent. Blanchard et al.
ritical Fluids 43 (2007) 150–180 169
[rCiftiiwiaaie
w[ploSteorbTTtttta[PtiCr
FaC
FaC
ttCaado
cwadmedegree of solubility in [bmim][PF6]. Solubilities of the solid
S. Keskin et al. / J. of Superc
125] demonstrated that CO2 is highly soluble in [bmim][PF6]eaching a mole fraction of 0.6 at 8 MPa. After extracting IL withO2 at 13.8 MPa and 40 ◦C there was no detectable [bmim][PF6]
n the extract, indicating that the solubility is less than 10−5 moleraction. In contrast to this result, a mixture of CO2 with conven-ional organic liquid results in significant solubility of the liquidn the CO2 phase. A mixture of 0.12 mole fraction naphthalenen [bmim][PF6] was extracted with CO2 at 13.8 MPa and 40 ◦Cith recoveries of 94–96% and therefore it was concluded that
t is possible to quantitatively extract an organic solute havingreasonable molecular weight from IL using scCO2 without
ny cross contamination. Furthermore, the dissolution of scCO2n IL was completely reversible and pure IL remained after thextraction of naphthalene and depressurization [125].
The experimental work of Blanchard and Brennecke [118],hich showed that a wide variety of solutes can be extracted from
bmim][PF6] with scCO2, with recovery rates greater than 95%,rovided a significant step to visualize the partnership estab-ished by the IL–CO2 system. One of the essential problemsf ILs, namely product recovery, was solved by applying theCFE technique. Hexane and benzene were chosen as the roots
o which numerous substituents groups were added to explore theffect of chemical structure on the solubility and extractabilityf an organic solute in and from an IL. The substituent groupsepresented were halogen, alcohol, ether, amide, ketone, car-oxylic acid, ester, and aldehyde with a wide range of polarity.he extraction experiments were conducted at 40 ◦C and 138 bar.he authors noted that for some organic solutes, scCO2 extrac-
ion achieved greater than 98% recoveries before the extractionest was terminated. These high recovery rates clearly indicatedhat although the ionic nature of IL might lead to an interac-ion with solute, it did not limit the extent of reaction. Benzenend chlorobenzene which exhibited phase immiscibility withbmim][PF6] required the least amount of CO2 for recovery.henols, benzoic acid and benzamide which are solids at room
emperature, required the largest amount of CO2. Figs. 10 and 11llustrate % recovery of solutes as a function of molar ratio ofO2 passed through the extractor to organic solute loaded in the
eactor (solute dipole moments are also given in these figures in
ig. 10. Extraction of aromatic solutes from [bmim][PF6] with scCO2 at 40 ◦Cnd 138 bar. (Reprinted with permission from [118]. Copyright 2001 Americanhemical Society)
ol
Fwr
ig. 11. Extraction of aliphatic solutes from [bmim][PF6] with scCO2 at 40 ◦Cnd 138 bar. (Reprinted with permission from [118]. Copyright 2001 Americanhemical Society)
erms of Debye). In this paper they showed that there was a rela-ion between the dipole moments of solutes and the amount ofO2 required for extraction, concluding that the organics withdipole moment of zero (such as benzene, hexane, cyclohex-
ne, etc.) are easily extracted compared to solutes with nonzeroipole moments. Fig. 12 illustrates the number of CO2 per molef organic solute as a function of dipole moment.
Studies on the solubilities of organics in [bmim][PF6] werearried out under ambient conditions, 22 ◦C and 0.98 bar [118],here IL–solute mixtures were stirred in closed containers to
void contamination with air and water vapor. Their resultsemonstrated that organics with the potential for strong inter-olecular interactions, those with a large dipole moment for
xample, generally exhibited complete miscibility or a large
rganic solutes were considerably less than the solubilities of theiquid organics, with the exception of the phenol. The authors
ig. 12. Effect of solute dipole moment on ease of extraction of [bmim][PF6]ith scCO2 at 40 ◦C and 138 bar. (Reprinted with permission from [118]. Copy-
ight 2001 American Chemical Society)
1 ritica
cicmnp
srsdtcacvwawfgdf
twaasm1s[
iid[npaif
tsapitstoneht
rC
cAelolnrrtcgceoKiNe(iw
sr
7
atIisf[sa[
tosIo
[mpfl
70 S. Keskin et al. / J. of Superc
oncluded that the solubilities of the benzene-based organ-cs were significantly greater than those of their hexane-basedounterparts. Solubility measurement results indicated that aro-atic compounds are more soluble in [bmim][PF6] than in
on-aromatic compounds of equivalent molecular weight andolarity.
Distribution coefficient (K), may be defined as the ratio ofolute mole fractions in the supercritical and in the IL phases,espectively. Since ILs do not dissolve in scCO2 appreciably,upercritical phase is essentially CO2 and organic solute. Theistribution coefficient is an important thermodynamic propertyo comprehend the IL–scCO2–solute systems. Theoretically, onean anticipate the trend of distribution coefficient between the ILnd the scCO2 phases by considering the volatility and polarityharacteristics of the solute. For example, a solute with a higholatility and low polarity will have a large affinity for CO2,hereas a solute with high polarity and aromaticity will havelarge affinity for the IL-rich liquid phase. Therefore, solutesith high polarities give a small K value due to high affinity
or IL and low affinity for CO2. Conversely, nonpolar solutesive large K values as a result of high affinity for CO2. With thisiscussion, it is clear that the compounds that have high affinityor CO2 can be more easily extracted from the IL mixture.
The phase behaviors of the organic solute–CO2 binary sys-ems also affect the ease of extraction of a compound from ILsith scCO2. Investigating the liquid phase data at low pressures,measure of affinity of compounds for CO2 was determined, andlso that compounds in which CO2 readily dissolves at low pres-ures have a greater attraction for CO2 [118]. Furthermore, theole ratio of CO2 to solute for 95% recovery is determined as
840 for CO2–cyclohexane binary system [149] however, theame ratio is as high as 20,300 for CO2–acetophenone system150].
Using scCO2 to recover products from IL is a good alternativef the products are thermally sensible or nonvolatile. A high boil-ng point solute can also be recovered from IL with scCO2, whereistillation is not a desirable option. Blanchard and Brennecke118] extracted a high-boiling point (230 ◦C) organic liquidamely 1,4-butanediol from [bmim][PF6]. This high-boilingoint solute followed the same solubility and extractability trends the other liquid solutes. Therefore, SCFE technology is read-ly applicable for the recovery of various kinds of compoundsrom ILs.
Scurto et al. [151] studied the use of scCO2 as a separa-ion switch for IL–organic mixtures. They demonstrated that theolutions of methanol and 3-butyl-1-methyl-imidazolium hex-fluorophosphate ([C4mim][PF6]) can be induced to form threehases in the presence of scCO2. Although the original solutions quite dilute in IL, application of scCO2 induces the forma-ion of an additional liquid phase which is rich in IL. This studyhowed that there is an alternative way that CO2 may be usedo separate ILs from organic compounds and the extraction ofrganic materials can be achieved without any IL cross contami-
ation in the recovered product. In their following study, Scurtot al. [152] demonstrated that separation of hydrophobic andydrophilic imidazolium-based ILs from aqueous solutions byhe application of scCO2 is possible and [bmim][PF6] is sepa-
i[It
l Fluids 43 (2007) 150–180
ated from an IL-saturated aqueous solution at 293 K and at aO2 pressure of 4.9 MPa.
The result of the mentioned studies presented that organicompound–IL–scCO2 mixture has a complex phase behavior.ccording to the results of their research Scurto et al. [151]
mphasized that during the IL–scCO2 reaction studies in whicharger amounts of organic reactant and products are present,ne must be aware of the possible formation of the additionaliquid phases that might contain only part of the componentsecessary for the desired reaction. After showing ILs could beecovered from methanol and water using CO2-induced sepa-ation, the same research group [153] investigated the factorshat control the vapor–liquid–liquid equilibrium in IL–organicompound–CO2 ternary systems via studying on several homo-eneous IL–organic compound mixtures. The experiments wereonducted at 40 ◦C. The results showed that the lowest criticalndpoint pressure (LCEP) was dependent on the choice of bothrganic, IL and the initial concentration of IL in the organic. The-point pressure was however independent of the type of IL, was
dentical with the organic compound–CO2mixture critical point.ajdanovic-Visak et al. [154] investigated the vapor–liquid
quilibrium of ternary (1-butanol–water–CO2) and quaternary[C4mim][NTf2]–1-butanol–water–CO2) systems. The demix-ng pressures of both mixtures were strongly controlled by theater concentration.The studies concerning the solute recovery from ILs by
cCO2, the system components, experimental conditions, majoresults of the studies and related references are given in Table 5.
.5. Other applications of IL–scCO2 systems
Brennecke and Maginn [8] discussed the potential industrialpplications of ILs in many areas such as catalytic reac-ions, liquid–liquid extractions, gas separations etc. After theL–scCO2 systems and the advantages of these systems are real-zed, a number of studies have been done for IL–CO2 biphasicystems: It is shown that a desired solute may be extractedrom an IL using scCO2 without any cross contamination118,125]. The use of scCO2 to separate ILs from their organicolvents [151]; the addition of CO2 to separate hydrophobicnd hydrophilic imidazolium-based ILs from aqueous solutions152] have all been important applications of IL–scCO2 systems.
Dzyuba and Bartsch [155] demonstrated the recent applica-ions of room temperature IL–scCO2 systems in metal catalyzedrganic reactions and enzyme-catalyzed transformations. Theolubility or stability of organometallic or enzymatic catalysts inLs and their negligible solubility in scCO2 is the basic advantagef IL–scCO2 systems.
Several groups have studied the IL–scCO2 reaction systems156–168]. In these studies, mostly IL was used as reactionedia and scCO2 was used as transport media for reactants and
roducts. Cole-Hamilton et al. [169] summarized the continuousow homogeneous catalysis using IL–scCO2 biphasic systems
n detail according reaction types. Also, Gordon and Leitner170] mentioned some of IL–scCO2 biphasic reaction systems.n the following paragraphs, IL–scCO2 biphasic reaction sys-ems are summarized.
S. Keskin et al. / J. of Supercritical Fluids 43 (2007) 150–180 171
Table 5Recovery of organic compounds from IL by scCO2
System Solute Temperature Pressure Findings Reference
[bmim][PF6]–CO2 Naphthalene 25–40 ◦C Up to 40 MPa CO2 is highly soluble in [bmim][PF6] [125]Two phases are not completely miscibleCO2 may be used to extract naphthalene, model solute, from an ILThe dissolution of CO2 in IL is completely reversible: pure IL remainsafter extraction of naphthalene and depressurizationCO2 can extract a wide variety of organic solutes from an ILIL contamination in the recovered product is eliminated by using CO2
All organic solutes exhibited recovery greater than 95%Intermolecular interactions between organic solutes and [bmim][PF6]have an effect on the solubility of solute in IL, but these interactions donot limit the degree of recovery
[bmim][PF6]–CO2 Benzene 40 ◦C 138 bar CO2 can extract a wide variety of organic solutes from an IL [118]Chlorobenzene IL contamination in the recovered product is eliminated by using CO2
Phenol All organic solutes exhibited recovery greater than 95%Anisole Intermolecular interactions between organic solutes and [bmim][PF6]
have an effect on the solubility of solute in IL, but these interactions donot limit the degree of recovery
Sellin et al. [156] and Webb et al. [157] studied the hydro-ormylation of alkenes in IL–scCO2 biphasic reaction mediand described the continuous flow homogeneous catalysis inL–scCO2 biphasic system. They dissolved the catalyst in ioniciquid and used scCO2 as the transport medium for the sub-trates and products. In Sellin et al. [156], they demonstratedhe hydroformylation of several alkenes such as 1-hexene, 1-onene and 1-octene in IL–scCO2 biphasic mixture. First, theydroformylation of 1-hexene was studied using triphenylphos-hite as the rhodium-based ligand ([Rh2(OAc)4]/P(OPh3)) inbmim][PF6] using scCO2. The hydroformylation of 1-hexeneas also performed in [bmim][PF6] without scCO2. The results
howed that addition of scCO2 to the reaction mixture lowershe conversion from >99–40%, but the selectivity and linearo branched (l:b) ratio were enhanced from 15.7 to 83.5% and.4 to 6.1, respectively. The success of this system encour-ged the authors to carry out similar reactions of 1-hexenend 1-nonene with repetitive uses of the same catalyst toearch for the possibility of continuous flow homogeneousatalysis. They found that the catalyst retained its activity andelectivity for only three to three runs, so in order to elimi-ate this drawback, the ligand–catalyst system was changed.
Ph2P(C6H4SO3)]–[bmim] was used together with [Rh2(OAc)4]s the catalyst precursor for the hydroformylation of 1-nonenen the [[bmim][PF6]–scCO2 biphasic system. The products wereushed from the reactor with scCO2. The activity of this cata-
h([e
yst system remained high for 12 runs with an acceptable l:batio. However, after the ninth run Rh leaching became impor-ant which was attributed to ligand oxidation. Contaminationith air during the many openings of the reactor may cause
he oxidation, which will be eliminated during continuous flow.inally, Sellin et al. [156] demonstrated the continuous hydro-ormylation of 1-octene using [PhP(C6H4SO3)2]–[pmim]2 andRh2(OAc)4] dissolved in [bmim][PF6]. The reactants andhe products were transported into and out of the reaction
edium via scCO2. The total pressure, temperature and reac-ion time were 200 bar, 100 ◦C and 33 h. The results showedhat the catalyst was stable at least between 8 and 10 h reac-ion time, no ligand oxidation occurred and the l:b ratiof the product aldehydes was 3.1. In another study, Webbt al. [157] investigated the hydroformylation 1-dodecenehich was representative for hydroformylation of relatively
ow volatility alkenes. The reactions were catalyzed by eitherh/[prmim][Ph2P(3-C6H4SO3)] or Rh/[prmim][TPPMS]. The
atter one was used for its easiness in crystallization and purifi-ation. They performed optimization reactions in order to obtainigher rates. They investigated the effect of the ionic liquid,ubstrate flow rate, temperature, gas composition on the rate of
ydroformylation of 1-octene. Among several ionic liquids used[bmim][PF6] [bmim][NTf2], [octmim][PF6], [octmim][NTf2],decmim][NTf2]), [octmim]NTf2 was found to be the mostffective in terms of high conversion rates attained (>80%).
1 ritica
Tw(anhwcf1b(abdup[Tfslt
b[eCo
o[sroartatosenCahtutpr[[([[
[Ewew(oo[8(sa[Itss
buphctu94ossToiabirhs2sl
bIahbtfaB
72 S. Keskin et al. / J. of Superc
he authors found that the catalyst turnover frequency (TOF),hich indicates productivity, could be >500 h−1 as the substrate
1-octene) flow rate was increased (>0.3 mL/min) at 200 barnd 100 ◦C. They concluded that continuous flow homoge-eous catalysis in IL–scCO2 biphasic system can be used forydroformylation of long chain alkenes at rates comparableith the ones found in commercial systems. In the latest study
onducted by Cole-Hamilton and co-workers [158] on hydro-ormylation of alkenes, they achieved rapid hydroformylation of-octene (rates up to 800 h−1) with the catalyst remaining sta-le for at least 40 h and with very low rhodium leaching levels0.5 ppm). A new system, in which the substrate, reacting gasesnd products dissolved in scCO2 and were flowing over a fixeded “supported ionic liquid phase (SILP)” catalyst, was intro-uced. The reactants in scCO2 (1-octene, CO and H2) flowedpwards through a tubular reactor containing a catalyst com-osed of [prmim][Ph2P(3-C6H4SO3)], [Rh(acac)(CO)2] andoctmim][Tf2N] supported on silica gel at 100 ◦C, and 100 bar.he authors claimed that this new system provided excellent dif-
usion of the substrate and gases to the catalyst surface, excellentolubility of the substrates and gases within the supported ioniciquid and extraction of heavy products that might otherwise foulhe catalyst by filling the pores.
There were two other studies on SILP catalysis [171,172],ut both studies were conducted in batch mode. Wang et al.171] described the synthesis of cyclic carbonates from CO2 andpoxides over silica-supported quaternary ammonium salts andiriminna et al. [172] studied the aerobic oxidation of alcoholsver a perruthenate catalyst.
Brown et al. [159] studied the asymmetric hydrogenationf tiglic acid catalyzed by Ru(O2CMe)2((R)–tolBINAP) inbmim]PF6 with addition of water as cosolvent. In this study,cCO2 was not involved in the reaction, but it was used toecover the products from the IL after the reaction time wasver. The results were evaluated in terms of enantioselectivitynd conversion. The catalyst/ionic liquid solution was reusedepeatedly (five cycles) without significant loss of enantioselec-ivity or conversion. In another work of the same group [160],number of different solvents were studied in order to evaluate
he most effective system for the enantioselective hydrogenationf �,�-unsaturated acids. Different solvent systems comprisedcCO2, different ILs, ILs with cosolvents and also CO2xpanded ionic liquids. They studied two types of substrates,amely class I (atropic acid) and class II (tiglic acid) substrates.lass I substrates were hydrogenated in high enantioselectivityt high H2 concentration whereas Class II substrates wereydrogenated in high enantioselectivity at low H2 concen-ration. For both substrates, Ru(O2CMe)2(R–tolBINAP) wassed as the catalyst. Atropic acid was hydrogenated withhe highest enantioselectivity (92%) in methanol (50 bar H2ressure). Lower enantioselectivity values were obtained foreactions in ILs ([bmim][PF6]—32%, [bmim][BF4]—15%,emim][O3SCF3]—25%, [emim][N(O2SCF3)2]—31%,
dmpim][N(O2SCF3)2]—39%) and IL–cosolvent systems[bmim][PF6]–toluene—19%, [bmim][PF6]–PrOH—33%,bmim][PF6]–MeOH—54%) at 50 bar H2 pressure. [bmim]PF6], [bmim][PF6]–toluene, [bmim][PF6]–PrOH, [bmim]
Arvr
l Fluids 43 (2007) 150–180
PF6]–MeOH systems were expanded by CO2 (50 bar CO2).xpansion caused increases in enantioselectivity consistentith greater H2 solubility and mass transfer rate in all mediums
xcept for [bmim][PF6]–toluene. Tiglic acid was hydrogenatedith the highest enantioselectivity (95%) in [emim]N(O2SCF3)2
5 bar H2). Reasonably high enantioselectivity values were alsobtained for other ionic liquids, as the authors expected, becausef low H2 concentration ([dmpim][N(O2SCF3)2]—93%,bmim][PF6]—93%, [mbpy][BF4]—88%, [bmim][BF4]—8%, [emim][O3SCF3]—84%). IL–cosolvent systems[bmim][PF6]–toluene, [bmim][PF6]–PrOH) were lesselective than hydrogenation in IL ([bmim][PF6]) alone. Tigliccid hydrogenations were also conducted in [bmim][PF6],bmim][PF6]–toluene, [bmim][PF6]–PrOH expanded systems.n these CO2 (70–bar) expanded systems, decreases in enan-ioselectivity were observed when compared to non-expandedystems. The decrease in enantioselectivity for [bmim][PF6]ystem was from 93% (non-expanded) to 85% (expanded).
Another example of hydrogenation reaction in IL–scCO2iphasic system was reported by Liu et al. [161]. IL phase wassed to immobilize the transition metal catalyst, and scCO2hase was used to recover the products. They examined theydrogenation of 1-decene and cyclohexene using Wilkinson’satalyst RhCl(PPh3)3, and hydrogenation of carbon dioxide inhe presence of dialkylamines to form N,N-dialkylformamidessing RuCl2(dppe)2 in [bmim][PF6]–scCO2 biphasic system.8% conversion was attained for hydrogenation of 1-decene at8 bar H2 and 207 bar total pressure (TOF: 410 h−1) at the endf 1 h reaction time. Hydrogenation of cyclohexene under theame experimental conditions mentioned above proceeded morelowly (82% conversion after 2 h, 96% conversion after 3 h).hey also performed four repetitive batch runs for hydrogenationf 1-decene, and demonstrated the efficient catalyst recycling viammobilization in ionic liquid. The hydrogenation of 1-decenend cyclohexene were also done with [bmim][PF6]–n-hexaneiphasic system. The results showed that there was no reactiv-ty advantage for CO2 over n-hexane for simple hydrogenationeactions. The conversion and selectivity were much higher forydrogenation of carbon dioxide in [bmim][PF6]–scCO2 bipha-ic medium starting with the di-n-propylformamide (80 ◦C,76 bar, 5 h), when they were compared to the conversion andelectivity values obtained only in scCO2 for less bulky diethy-amine and n-propylamine.
Bossmann et al. [162] also utilized the benefits of IL/CO2iphasic system to immobilize the organometallic catalyst inL phase and recover the product without harming the cat-lyst. They investigated the continuous flow system for theydrovinylation of styrene with Wilke’s catalyst in IL/CO2iphasic system. They initially identified the suitable ILshat would allow the activation of the catalyst. It wasound that the activation strongly depended on the IL’snion type. The reaction conversion rates for the anionsARF (BARF: tetrakis[3,5-bis(trifluoromethyl)phenyl]borate),
l[OC(CF3)2Ph]4, Tf2N and BF4 were 100, 90.5, 69.9 and 39.6,
espectively for [emim] cation. It was also noted that the acti-ation of the catalyst was not just a simple anion-exchangeeaction, and the specific environment of the ionic solvent sys-
ritica
t(cbwiwwdwlattbitvug
mwwswsastTlsatwasgcaba(cooo[1ohae(is
cta[sr
csapbacaspeTa1[aidt(1dttlttrbL11daiehet11soto
S. Keskin et al. / J. of Superc
em seemed to activate the catalyst. When [emim] and [4-mbp]4-mbp: 1-n-butyl-4-methylpyridinium) were used as differentations with the same anion, it was reported that at compara-le conversions, higher enantioselectivity values were foundith [4-mbp][BF4] and [4-mbp][Tf2N] than the correspond-
ng [emim] salts. The continuous flow styrene hydrovinylationas conducted in [emim][Tf2N]–CO2 at 80 bar. The catalystas stable over a reaction time of 61 h and enantioslectivityecreased slightly at that period, while products were extractedith compressed CO2. Tkatchenko et al. [163] studied the pal-
adium catalyzed dimerization of methyl methacrylate at 83 ◦Cnd 200 bar in [bmim][BF4]–scCO2 biphasic system. The selec-ivity (>98%) was identical to that of the monobasic system. Theurnover frequency and turnover numbers were comparable foroth monophasic and biphasic systems, increasing with increas-ng substrate to palladium ratios in the IL phase. They concludedhat, as the CO2-rich phase acted as a substrate and product reser-oir, there was a possibility for the reaction to be conductednder continuous feed and product recovery conditions withreener solvents.
Hou et al. [164] investigated the oxidation of 1-hexene byolecular oxygen in [bmim][BF6]–scCO2 biphasic system asell as in [bmim][BF6], scCO2, and in the absence of solventith catalysts PdCl2 and CuCl2. The results showed that the
electivity to the desired product 2-hexanone was much higherhen the reaction was done in [bmim][BF6]–scCO2 biphasic
ystem (125 bar, 333.2 K and 17 h of reaction time). Addition-lly, the catalyst was more stable in biphasic system than it was incCO2 only. Kawanami et al. [165] performed the chemical fixa-ion of CO2 to cyclic carbonates in a IL–scCO2 biphasic system.hey reported that synthesis of propylene carbonate from propy-
ene oxide and carbon dioxide in [omim][BF4]–scCO2 biphasicystem (14 MPa, 100 ◦C) was achieved with nearly 100% yieldnd selectivity within 5 min and TOF value was 77 times higherhan those so far reported. They also observed that the yieldas increased as the alkyl chain length of the IL’s cation was
lso increased from C2 to C8. Moreover, different epoxide sub-trates having phenyl substituted groups and alkyl side chainroups were examined for the synthesis of the correspondingarbonates in [omim][BF4]–scCO2 biphasic system at 14 MPand 100 ◦C. Gao et al. [166] studied the transesterificationetween isoamyl acetate and ethanol in scCO2, [bmim][BF6]nd [bmim][BF6]–scCO2 system using p-xylenesulfonic acidp-TSA) as the catalyst. The results showed that the equilibriumonversion in [bmim][BF6]–scCO2 media was lower than thosebserved in scCO2 or [bmim][BF6]. An interesting applicationf IL–scCO2 biphasic system in synthesis was electro-oxidationf benzyl alcohol to benzaldehyde in an electrochemical cell167]. The reaction was carried out at 318.2 K and up to0.3 MPa, and two ILs, 1-butyl-3-methylimidazolium tetraflu-roborate ([bmim][BF4]) and 1-butyl-3-methylimidazoliumexafluorophosphate ([bmim][PF6]), were used as the solventsnd electrolytes. [bmim][BF4] was a better medium for the
lectro-oxidation of benzyl alcohol. The Faradic efficiencyFE) and selectivity of benzaldehyde increased as the pressurencreased up to 9.3 MPa, whereas the FE decreased as the pres-ure was increased further. The authors noted that the products
brcb
l Fluids 43 (2007) 150–180 173
ould be easily recovered from the IL by using scCO2 extrac-ion after the electrolysis, and the IL could be reused. Yoon etl. [168] studied Heck coupling of iodobenzene with olefins inbmim][PF6] catalyzed by PdCl2/Et3N. They did not incorporatecCO2 in the reaction, it was used after the reaction for productecovery.
IL–scCO2 biphasic systems were also used for enzymeatalysis [173–179]. Lozano et al. [173–174] investigated theynthesis of butyl butyrate from vinyl butyrate and 1-butanol,nd the kinetic resolution of rac-1-phenylethanol with vinylropionate by transesterification. They used both free and immo-ilized C. antarctica lipase B (CALB) in IL ([emim][Tf2N]nd [bmim][Tf2N]) as catalyst for continuous biphasic bio-atalysis. CO2 was used as transport medium for reactantsnd products. The synthetic activity of the enzyme in IL–CO2ystem was tested through operation/storage cycles. Operationeriod (4 h) was followed by a storage period (20 h) of thenzyme–IL system under dry conditions at room temperature.he continuous synthesis of butyl butyrate from vinyl butyratend 1-butanol by transesterification was studied at 12.5 and5 MPa at 40, 50 and 100 ◦C in [bmim][Tf2N]–CO2 system173]. The specific activity and the selectivity were enhanceds the temperature increased, the selectivity was high (>95%)n all cases giving higher than 50% conversions. The activityecay for the repetitive uses of the free enzyme–IL system washe highest at high temperature. The continuous synthesis ofR)-1-phenylethyl propionate from the kinetic resolution of rac--phenylethanol with vinyl propionate catalyzed by free CALBissolved in [emim][Tf2N] and [bmim][Tf2N] was also inves-igated at 15 MPa, 50 and 100 ◦C [173]. The results showedhat the selectivity of the reaction, the activity and the half-ife of the enzyme–IL system were lower for this reaction thanhose observed for butyl butyrate synthesis. However, high enan-ioselectivity (>99.9%) was exhibited by the enzyme. As theeaction temperature increased, the selectivity was increasedut the specific activity decreased. In another publication ofozano et al. [174], the continuous kinetic resolution of rac--phenylethanol in IL–CO2 biphasic system was investigated at20 and 150 ◦C and 10 MPa. Both free and immobilized CALBispersed in [emim][Tf2N] and [bmim][Tf2N] were used as cat-lyst. [emim][Tf2N] was a better IL in terms of obtaining highernitial synthetic activity and longer half-life time of the freenzyme–IL system. Immobilized enzyme–IL system had longeralf-life time and higher synthetic activity compared to freenzyme–IL system. No activity losses during successive opera-ion cycles were observed for immobilized enzyme at 120 ◦C and0 MPa. Additionally, at the same operating conditions (120 ◦C,0 MPa), in immobilized CALB–[emim][Tf2N] system, highelectivity values (>98%) were always obtained for successiveperation cycles, whereas in free CALB–[emim][Tf2N] system,he selectivity increased from 36 to 98.5% as the number ofperation cycles increased.
Reetz et al. [175] studied the acylation of 1-octanol
y vinyl acetate (batch and continuous mode) and kineticesolution of 1-phenylethanol (batch mode), which wereatalyzed by CALB in [bmim][BTA]–CO2 system (BTA:is(trifluoromethanesulfonamide)). The acylation of 1-octanol
1 ritica
baetusae1lCfmfatatlIshripit6ttTirlrptA1ae
fom1[flstibogeud
ti
sl[eociuLmraeemc[t(otaC[otft
pmistir9g
siwcmntttdt
74 S. Keskin et al. / J. of Superc
y vinyl acetate in continuous mode was performed at 10.5 MPand 45 ◦C. A total yield of 93.9% was obtained after 24 h. Reetzt al. [180] introduced a new method for enantiomer separa-ion utilizing biocatalytic kinetic resolution and SCF extractionsing an IL–scCO2 system. Application of this new method foreparation of racemic secondary alcohols by CALB catalysisnd CO2 extraction was represented. They converted sev-ral alcohols (2-octanol, 1-phenyl-ethanol, 3-methyl-2-butanol,-(2-phenylethyl) ethanol, and 1-(2-napthyl) ethanol) enantiose-ectively to corresponding acetates and laureates by immobilizedALB suspended in [bmim][BTA], and separate (S)-alcohol
rom the product (R)-ester via CO2. Vinyl laureate was found toake the ester less soluble in CO2 than the alcohol, which allows
or efficient separation, therefore it was used as the acylationgent. In that work, both batch and continuous modes of opera-ion were studied. In batch mode, when vinyl laureate was useds the acylation agent, in the early fractions of isobaric extrac-ion, the alcohol was extracted with high selectivity, whereas theauryl ester was obtained with high purity in the later fractions.n order to speed up the extraction for the ester-rich fraction, theyuggested increasing the CO2 density after most of the alcoholas been isolated. This procedure was successful for the sepa-ation of several alcohols and lauryl esters. Additionally, theynclude a separation chamber leading to a two-step extractionrocedure. First, extraction was done at 60 ◦C and 105 bar, thenn the separation chamber the pressure was reduced to 80 bar andhen the CO2 was vented through cryo-traps. By this procedure6% of the theoretical amount of the alcohol under investiga-ion was isolated with an enantiomeric purity >99.9% and lesshan 0.5% contamination with its corresponding lauryl esters.hen the pressure of reactor and the extraction chamber was
ncreased to 200 bar and 89% of the theoretical amount of lau-yl ester was isolated with an enantiomeric purity >99.9% andess than 1% contamination with alcohol. Kinetic resolution ofac-1-phenylethanol was used to demonstrate the continuousrocess. The separation after extraction was provided throughwo separation chambers with two steps of density reduction.fter 112 h operation, 81% of the theoretical amount of the (S)--phenylethanol was isolated with an enantiomeric purity >97%nd less than 0.1% contamination with its corresponding laurylster.
Lozano et al. [176] studied the synthesis of glycidyl estersrom rac-glycidol catalyzed with free and immobilized formsf lipases from C. antarctica (CALA and CALB) and Mucoriehei (MML) in toluene, IL and IL–scCO2 (40, 50 ◦C and00, 150 bar). Four different ILs were used: [emim][NTf2],bmim][PF6], [bmim][NTf2], and trioctylmethylammonium tri-imide ([troma][Tf2N]). Using ILs instead of a classical organicolvent (toluene) and the increase in the alkyl chain length ofhe acylation agent had both positive effect on the enzyme activ-ty and when these effects combined, the synthetic activity cane enhanced 95 times. CALA and MML favored the synthesisf R-glycidyl esters, whereas CALB favored the synthesis of S-
lycidyl esters. CALB showed the highest activity among othernzymes. CO2 was used to transport the substrates and prod-cts. The synthetic activity of the free and immobilized lipasesecreased in IL–scCO2 biphasic system, but the enantioselec-
hfia
l Fluids 43 (2007) 150–180
ivity remained unchanged with respect to the values obtainedn only IL reaction media.
CALB catalyzed ester synthesis in IL–scCO2 biphasicystems with five different ILs ((3-hydroxypropyl)-trimethy-ammonium bis(trifluoromethylsulfonyl) imide [C3OHtma]NTf2]; (3-cyanopropyl)-trimethylammonium bis(trifluorom-thylsulfonyl) imide [C3CNtma][NTf2]; butyl-trimethylamm-nium bis(trifluoromethylsulfonyl) imide, [C4tma][NTf2]; (5-yanopentyl)-trimethylammonium bis(trifluoromethylsulfonyl)mide [C5CNtma][NTf2]; hexyl-trimethylammonium bis(trifl-oromethylsulfonyl) imide, [C6tma][NTf2]) were studied byozano et al. [177]. The suitability of these ILs as enzy-atic reaction media was tested for the kinetic resolution of
ac-phenylethanol by transesterification with vinyl propionate,nd all of the tested ILs were found to be suitable media fornzyme catalysis. The synthetic activities and stabilities of thenzyme were determined in these five ILs. Then, the perfor-ance of CALB catalysis in IL–scCO2 biphasic system in
ontinuous operation was tested for [C4tma][NTf2]–scCO2 andC3CNtma][NTf2]–scCO2 systems at 50 ◦C and 10 MPa forhe synthesis of several short-chain alkyl esters (butyl acetateBA), butyl propionate (BP), butyl butyrate (BB), hexyl propi-nate (HP), hexyl butyrate (HB), and octyl propionate (OP)), byransesterification from the respective vinyl alkyl esters withlkyl-1-ols. The results showed that the synthetic activity ofALB in [C4tma][NTf2]–scCO2 system was higher than that in
C3CNtma][NTf2]–scCO2, even though the opposite result wasbtained for activity values in pure IL. The authors concludedhat rate-limiting parameters (synthetic activity and mass trans-er phenomena between IL and scCO2 phases) were related withhe solubility parameter of the IL’s alkyl chain and reagents.
Hernandez et al. [178] presented the synthesis of butylropionate in scCO2 and IL–scCO2 using a recirculating enzy-atic membrane reactor, in which �-alumina membranes were
mmobilized with CALB. The reactor was tested in onlycCO2 at 50 ◦C and 80 bar. In the second part of the study,he immobilized enzyme was coated with three different ILs,.e. [bmim][PF6], 1-butyl-2,3-dimethylimidazolium hexafluo-ophosphate [bdimim][PF6], [omim][PF6]. The selectivity was5% when only scCO2 medium was utilized, and it increased toreater than 99.5% when IL–scCO2 biphasic system was used.
Lozano et al. [179] investigated the importance of theupporting material for the activity of immobilized CALBn IL–hexane and IL–scCO2 biphasic systems at controlledater activity. For this purpose they immobilized a commer-
ial solution of free CALB onto twelve different silica supports,odified with specific side chains (e.g. alkyl, amino, carboxylic,
itrile, etc.) by adsorption. Both biphasic media was tested forhe kinetic resolution of rac-phenylethanol by transesterifica-ion with vinyl propionate. CALB activity and stability wasested IL–hexane and IL–scCO2 biphasic systems with twoifferent ILs: [btma][NTf2] and [toma][NTf2] (btma: butyl-rimethylammonium, toma: trioctyl-methylammonium). The
ighest synthetic activity of immobilized enzyme was obtainedor butyl-silica derivative, however for all supports, the selectiv-ty for the reaction was higher than 94% except for the quaternarymmonium-Si support and the enantiomeric excess was greater
S. Keskin et al. / J. of Supercritical Fluids 43 (2007) 150–180 175
aning
tIwip
aimlsodcrlttnf
8
pgbsTabaahs
a
Fig. 13. A process suggested for cle
han 99.9%. The immobilized activity decreased 10 times inL–hexane compared to that in only hexane, but half-life timesere enhanced by up to six times. The synthetic activity of
mmobilized CALB increased by six times in IL–scCO2 com-ared to that in hexane.
Recently, the potential of ILs to dissolve soil contaminants atmbient conditions and the ability of supercritical carbon diox-de (scCO2) to recover these contaminants from IL extracts were
utually utilized to clean contaminated soils [181]. Naphtha-ene was used as the model component to represent a group ofoil contaminants, and 1-n-butyl 3-methylimidazolium hexaflu-rophosphate ([bmim][PF6]) was used as the IL. The resultsemonstrated that soil contaminated with naphthalene wasleaned using [bmim][PF6], and the amount of naphthaleneemaining in the soil was below the allowable contaminationimit. This study was the first in the literature which investigates
he soil/model-contaminant/IL/scCO2 system. On the basis ofhe findings a process flowsheet for the IL extraction of contami-ated soils and continuous scCO2 extraction of the contaminantsrom IL extracts was suggested and is given in Fig. 13.15
atit
of soils using IL and scCO2 [181].
. Summary
This work aims to summarize and discuss the informationresent in the literature about ILs and IL–CO2 systems. Areat number of researchers investigated the high-pressure phaseehavior of IL–CO2 systems and concluded that CO2 is highlyoluble in most ILs, and ILs are not measurably soluble in scCO2.he effects of pressure, temperature, nature of the anion and thelkyl chain length of the cation on the solubility of CO2 reportedy various studies are discussed and summarized here. Volatilend nonpolar scCO2 has become a good partner of nonvolatilend polar IL and this new system with its unique propertiesave been utilized to extract organic compounds from ILs usingcCO2.
ILs are receiving more and more attention every day both incademic research and commercial applications and they seem
s good replacements for volatile organic solvents. However,here is a discussion about the greenness of the ILs due to theirncomplete physical, chemical and toxicological data. Although,here are some question marks related to the specific character-
1 ritica
itcpcmfv
A
R(taai
R
76 S. Keskin et al. / J. of Superc
stics of ILs, it seems that most of the researchers will continueo work with this new solvent, the developments of new appli-ations utilizing ILs will increase rapidly and the number ofublications will rise exponentially in the future. The presentosts of the ILs are quite prohibitive in many probable com-ercial applications. However, there are hopes that in the near
uture, the cost/benefit figures of the ILs will bring economiciability to their more common use.
cknowledgements
The financial supports provided by Bogazici Universityesearch Foundation via the Project No. 06A502 and TUBITAK
The Scientific and Technical Research Council of Turkey) viahe Project No. 104M185 are gratefully acknowledged. Theuthors thank the three anonymous reviewers for their detailednd insightful comments and suggestions, which significantlymproved the manuscript.
eferences
[1] T. Welton, Room temperature ionic liquids: Solvents for synthesis andcatalysis, Chem. Rev. 99 (1999) 2071–2084.
[2] P.L. Short, Out of the ivory tower, Chem. Eng. News 84 (2006)15–21.
[3] S. Sugden, H. Wilkins, The parachor and chemical constitution. Part.Fused metals and salts, J. Chem. Soc. (1929) 1291–1298.
[4] J. Gorman, Faster, better, cleaner? New liquids take aim at old-fashionedchemistry, Sci. News 160 (2001) 156–158.
[5] J.S. Wilkes, J.A. Levisky, R.A. Wilson, C.L. Hussey, Dialkylimidazoliumchloroaluminate melts, a new class of room-temperature ionic liquids forelectrochemistry, spectroscopy and synthesis, Inorg. Chem. 21 (1982)1263–1264.
[12] C. Lagrost, D. Carrieı, M. Vaultier, P. Hapiot, Reactivities of some elec-trogenerated organic cation radicals in room-temperature ionic liquids:toward an alternative to volatile organic solvents? J. Phys. Chem. A 107(2003) 745–752.
[13] A. Shariati, C.J. Peters, High pressure phase equilibria of systems withionic liquids, J. Supercrit. Fluids 34 (2005) 171–176.
[14] A. Shariati, K. Gutkowski, C.J. Peters, Comparison of the phase behaviorof some selected binary systems with ionic liquids, AIChE J. 51 (2005)1532–1540.
[15] H. Zhao, S. Xia, P. Ma, Review: use of ionic liquids as green solvents for
extractions, J. Chem. Technol. Biotechnol. 80 (2005) 1089–1096.
[16] G.J. Kabo, A.V. Blokhin, A. Paulechka, U. Ya, A.G. Kabo, M.P.Shymanovich, J.V. Magee, Thermodynamic properties of 1-butyl-3-methylimidazolium hexafluorophosphate in the condensed state, J. Chem.Eng. Data 49 (2004) 453–461.
l Fluids 43 (2007) 150–180
[17] J.G. Huddleston, H.D. Willauer, R.P. Swatloski, A.E. Visser, R.D. Rogers,Room temperature ionic liquids as novel media for clean liquid–liquidextraction, Chem. Commun. (1998) 1765–1766.
[18] C. Chiappe, D. Pieraccini, Review commentary; ionic liquids: sol-vent properties and organic reactivity, J. Phys. Org. Chem. 18 (2005)275–297.
[19] D.C. Donata, F. Marida, H. Migen, University of Torino, http://lem.ch.unito.it/didattica/infochimica/Liquidi%20Ionici/Definition.html(accessed 20 June, 2006).
[20] M.J. Earle, J.M.S.S. Esperanca, M.A. Gilea, J.N.C. Lopes, L.P.N. Rebelo,J.W. Magee, K.R. Seddon, J.A. Widegren, The distillation and volatilityof ionic liquids, Nature 72 (2006) 1391–1398.
[21] J. Dupont, C.S. Consorti, J. Spencer, Room-temperature molten salts:neoteric green solvents for chemical reactions and processes, J. Braz.Chem. Soc. 11 (2000) 337–344.
[22] F. Mutelet, V. Butet, J.N. Jaubert, Application of inverse gas chromatog-raphy and regular solution theory for characterization of ionic liquids,Ind. Eng. Chem. Res. 44 (2005) 4120–4127.
[23] P. Kolle, R. Dronskowski, Synthesis, crystal structures and electrical con-ductivities of the ionic liquid compounds butyl dimethylimidazoliumtetrafluoroborate, hexafluoroborate and hexafluoroantimonate, Eur. J.Inorg. Chem. (2004) 2313–2320.
[24] A.J. Carmichael, K.R. Seddon, Polarity study of some 1-alkyl-3-methylimidazolium ambient-temperature ionic liquids with thesolvatochromic dye, Nile red, J. Phys. Org. Chem. 13 (2000) 591–595.
[25] S.N.V.K. Aki, J.F. Brennecke, A. Samanta, How polar are room temper-ature ionic liquids? Chem. Commun. (2001) 413–414.
[26] R.A. Sheldon, R.M. Lau, F. Sorgedrager, K. van Rantwijk, R. Seddon,Biocatalysis in ionic liquids, Green Chem. 4 (2002) 147–151.
[27] B. Jastorff, R. Stormann, J. Ranke, K. Molter, F. Stock, B. Oberheitmann,W. Hoffmann, J. Hoffmann, M. Nuchter, B. Ondruschka, J. Filser, Howsustainable are ionic liquids? Structure–activity relationships and bio-logical testing as important elements for sustainability evaluation, GreenChem. 5 (2003) 136–142.
[31] L. Cammarata, S.G. Kazarian, P.A. Salter, T. Welton, Molecular states ofwater in room temperature ionic liquids, Phys. Chem. Chem. Phys. 23(2001) 5192–5200.
[34] A.S. Wells, V.T. Coombe, On the freshwater ecotoxicity and biodegrada-tion properties of some common ionic liquids, Org. Process Res. Dev. 10(2006) 794–798.
[35] R.P. Swatloski, J.D. Holbrey, S.B. Memon, G.A. Caldwell, K.A. Caldwell,R.D. Rogers, Using Caenorhabditis elegans to probe toxicity of 1-alkyl-3-methylimidazolium chloride based ionic liquids, Chem. Commun. (2004)668–669.
[36] R.J. Bernott, E.E. Kennedy, G.A. Lamberti, Effects of ionic liquids on thesurvival, movement, and feeding behavior of the freshwater snail, Physaacuta, Environ. Toxicol. Chem. 24 (2005) 1759–1765.
[37] T.D. Landry, K. Brooks, D. Poche, M. Woolhiser, Acute toxicity profile of1-butyl-3-methylimidazolium chloride, Bull. Environ. Contam. Toxicol.74 (2005) 559–565.
[38] C. Pretti, C. Chiappe, D. Pieraccini, M. Gregori, F. Abramo, G. Monnia,L. Intorre, Acute toxicity of ionic liquids to the zebra fish (Danio rerio),
Green Chem. 8 (2006) 238–240.
[39] P. Stepnowski, A. Muller, P. Behrend, J. Ranke, J. Hoffmann, B. Jastorff,Reversed-phase liquid chromatographic method fort the determinationof selected room-temperature ionic liquid cations, J. Chromatogr. A 993(2003) 173–178.
[40] N. Gathergood, M.T. Garcia, P.J. Scammells, Biodegradable ionic liquids.Part I. Concept, preliminary targets and evaluation, Green Chem. 6 (2004)166–175.
[41] N. Gathergood, P.J. Scammells, Design and preparation of room-temperature ionic liquids containing biodegradable side chains, Aust. J.Chem. 55 (2002) 557–560.
[42] E.B. Carter, S.L. Culver, P.A. Fox, R.D. Goode, I. Ntai, M.D. Tickell,R.K. Traylor, N.W. Hoffman, J.H. Davis Jr., Sweet success: ionic liquidsderived from non-nutritive sweeteners, Chem. Commun. (2004) 630–631.
[43] M.T. Garcia, N. Gathergood, P.J. Scammells, Biodegradable ionic liquids.Part II. Effect of the anion and toxicology, Green. Chem. 7 (2005) 9–14.
[44] G.-H. Tao, L. He, N. Sun, Y. Kou, New generation ionic liquids: cationsderived from amino acids, Chem. Commun. (2005) 3562–3564.
[45] N. Gathergood, P.J. Scammells, M.T. Garcia, Biodegradable ionic liquids.Part III. The first readily biodegradable ionic liquids, Green Chem. 8(2006) 156–160.
[46] G.-H. Tao, L. He, W.-S. Liu, L. Xu, W. Xiong, T. Wang, Y. Kou, Prepa-ration, characterization and application of amino acid-based green ionicliquids, Green Chem. 8 (2006) 639–646.
[47] P. Wasserscheid, R. van Hal, A. Bosmann, 1-n-Butyl-3-methylimi-dazolium ([bmim]) octylsulfate—an even ‘greener’ ionic liquid, GreenChem. 4 (2002) 400–404.
[48] C.D. Tran, S.H.D.P. Lacerda, D. Oliveira, Absorption of water by roomtemperature ILs: effect of anions on concentration and state of water, Soc.Appl. Spectrosc. 57 (2003) 152–157.
[49] J.G. Huddleston, A.E. Visser, W.M. Reichert, H.D. Willauer, G.A. Bro-ker, R.D. Rogers, Characterization and comparison of hydrophilic andhydrophobic room temperature ionic liquids incorporating the imida-zolium cation, Green Chem. 3 (2001) 156–164.
[50] D.C. Donata, F. Marida, H. Migen, University of Torino, http://lem.ch.unito.it/didattica/infochimica/Liquidi%20Ionici/Composition.html(accessed 20 June, 2006).
[51] S.K. Poole, P.H. Shetty, C.F. Poole, Chromatographic and spectroscopicstudies of the solvent properties of a new series of room-temperature liquidtetraalkylammonium sulfonates, Anal. Chim. Acta 218 (1989) 241–264.
[52] J.S. Wilkes, M.J. Zaworotko, Air and water stable 1-ethyl-3-methylimidazolium based ionic liquids, J. Chem. Soc. Chem. Commun.(1992) 965–967.
[54] J.D. Holbrey, W.M. Reichert, R.P. Swatloski, G.A. Broker, W.R. Pitner,K.R. Seddon, R.D. Rogers, Efficient, halide free synthesis of new lowcost ionic liquids: 1,3-dialkylimidazolium salts containing methyl- andethyl-sulfate anions, Green Chem. 4 (2002) 407–413.
[55] Y.R. Mirzaei, B. Twamley, J.M. Shreeve, Syntheses of 1-alkyl-1,2,4-triazoles and the formation of quaternary 1-alkyl-4-polyfluoroalkyl-1,2,4-triazolium salts leading to ionic liquids, J. Org. Chem. 67 (2002)9340–9345.
[57] A.E. Visser, R.P. Swatloski, W.M. Reischert, R. Mayton, S. Sheff, A.Wierzbicki, J.H. Davis, R.D. Rogers, Task-specific ionic liquids incorpo-rating novel cations for the coordination and extraction of Hg2+ and Cd2+:synthesis, characterization, and extraction studies, Environ. Sci. Technol.36 (2002) 2523–2529.
[58] R.A. Sheldon, in: M.P.C. Weijnen, A.A.H. Drienkenburg (Eds.), Preci-sion Process Technology: Perspectives for Pollution Prevention, Kluwer,Dordrecht, 1993, p. 125.
[59] A.G. Fadeev, M.M. Meagher, Opportunities for ionic liquids in recoveryof biofuels, Chem. Commun. (2001) 295–296.
[60] M.S. Selvan, M.D. McKinley, R.H. Dubois, J.L. Atwood, Liquid–liquidequilibria for toluene + heptane + 1-ethyl-3-methylimidazolium triiodideand toluene + heptane + 1-butyl-3-methylimidazolium triiodide, J. Chem.
Eng. Data 45 (2000) 841–845.
[61] T.M. Letcher, N. Deenadayalu, B. Soko, D. Ramjugernath, P.K.Naicker, Ternary liquid–liquid equilibria for mixtures of 1-methyl-3-octylimidazolium chloride + an alkanol + an alkane at 298.2 K and 1 bar,J. Chem. Eng. Data 48 (2003) 904–907.
l Fluids 43 (2007) 150–180 177
[62] W. Wu, K.N. Marsh, A.V. Deev, J.A. Boxall, Liquid–liquid equilibria ofroom temperature ionic liquids and butanol, J. Chem. Eng. Data 48 (2003)486–491.
[63] M. Wagner, O. Stanga, W. Schroer, Corresponding states analysis of thecritical points in binary solutions of room temperature ionic liquids, Phys.Chem. Chem. Phys. 5 (2003) 3943–3950.
[64] U. Domanska, A. Marciniak, Solubility of 1-alkyl-3-methylimidazoliumhexafluorophosphate in hydrocarbons, J. Chem. Eng. Data 48 (2003)451–456.
[65] A. Arce, A. Marchiaro, O. Rodriguez, A. Soto, Essential oil terpenlessby extraction using organic solvents or ionic liquids, AIChE J. 52 (2006)2089–2097.
[66] D. Camper, P. Scovazzo, C. Koval, R. Noble, Gas solubilities in room-temperature ionic liquids, Ind. Eng. Chem. Res. 43 (2004) 3049–3054.
[67] P. Scovazzo, D. Camper, J. Kieft, J. Poshusta, C. Koval, R. Noble, Regularsolution theory and CO2 gas solubility in room-temperature ionic liquids,Ind. Eng. Chem. Res. 43 (2004) 6855–6860.
[68] R.E. Baltus, B.H. Culbertson, S. Dai, H.M. Luo, D.W. DePaoli, Low-pressure solubility of carbon dioxide in room-temperature ionic liquidsmeasured with a quartz crystal microbalance, J. Phys. Chem. B 108 (2004)721–727.
[69] J.L. Anthony, E.J. Maginn, J.F. Brennecke, Solubilities and ther-modynamic properties of gases in the ionic liquid 1-n-butyl-3-methylimidazolium hexafluorophosphate, J. Phys. Chem. B 106 (2002)7315–7320.
[70] L.A. Blanchard, Z. Gu, J.F. Brennecke, High pressure phase behav-ior of ionic liquid/CO2 systems, J. Phys. Chem. B 105 (2001) 2437–2444.
[71] J.L. Anthony, E.J. Maginn, J.F. Brennecke, Solution thermodynamics ofimidazolium-based ionic liquids and water, J. Phys. Chem. B 105 (2001)10942–10949.
[72] J. Kumelan, A.P.-S. Kamps, D. Tuma, G. Maurer, Solubility of H2 in theionic liquid [bmim][PF6], J. Chem. Eng. Data 51 (2006) 11–14.
[73] S. Dai, D. DePaoli, M. Dietz, J. Mays, J. McFarlane, W. Steele, Technicalsummaries on ionic liquids in chemical processing, Chemical IndustryVision 2020 Technology Partnership Workshop, 2003.
[74] Air Products and Chemicals Inc., www.airproducts.com/NR/rdonlyres/9D919AAD-DE4F-4614-B3C4-E923C6BF786E/0/GasguardSAS.pdf,2005.
[75] J. Dupont, G.S. Fonseca, A.P. Umpierre, P.F.P. Fichtner, S.R. Teixeira,Transition-metal nanoparticles in imidazolium ionic liquids: recycablecatalysts for biphasic hydrogenation reactions, J. Am. Chem. Soc. 124(2002) 4228–4229.
[76] R. Morland, American Chemical Society Division of Industrial and Engi-neering Chemistry, in: Proceedings of the 221st American ChemicalSociety National Meeting, San Diego, 2001.
[78] P. Wasserscheid, Potential to apply ionic liquids in industry: exemplifiedfor the use as solvents in industrial applications as homogenous catal-ysis, in: Green Industrial Applications of Ionic Liquids, NATO ScienceSeries II: Mathematics, Physics and Chemistry 92, Kluwer AcademicPublishers, Dordecht, 2003, pp. 29–47.
[79] E. Ota, Some aromatic reactions using AlCl3-rich molten salts., J. Elec-trochem. Soc. 134 (1987), C512.
[80] T. Nemeth, L. Bricker, C.J. Holmgren, S.J. Monson, E. Lyle, Direct car-bonylation of paraffins using an ionic liquid catalyst, US Patent 628828(2001).
[81] M.J. Earle, S.P. Katdare, Oxidative halogenation of aromatic compoundsin the presence of an ionic liquid, World Patent WO 2002030852 (2002).
[82] M.J. Earle, S.P. Katdare, Oxidation of alkylaromatics in the presence ofionic liquids, World Patent WO 2002030862 (2002).
[91] P.A.Z. Suarez, J.E.L. Dullis, S. Einloft, R.F. de Souza, J. Dupont, The useof ionic liquids in two phase catalytic hydrogenation reaction by rhodiumcomplexes, Polyhedron 15 (1996) 1217–1219.
[92] P.A.Z. Suarez, J.E.L. Dullis, S. Einloft, R.F. de Souza, J. Dupont, Two-phase catalytic hydrogenation of olefins by Ru(II) and Co(II) complexesdissolved in 1-n-butyl-3-methylimidazolium tetrafluoroborate ionic liq-uid, Inorg. Chim. Acta 255 (1997) 207–209.
[93] Q. Liu, M.H.A. Janssen, F. van Rantwijk, R.A. Sheldon, Room temper-ature ionic liquids that dissolve carbohydrates in high concentrations,Green Chem. 7 (2005) 39–42.
[94] M. Matsumoto, K. Mochiduki, K. Fukunishi, K. Kondo, Extraction oforganic acids using imidazolium-based ionic liquids and their toxicity toLactobacillus rhamnosus, Sep. Purif. Technol. 40 (2004) 97–101.
[95] R.M. Lau, F. van Rantwijk, K.R. Seddon, R.A. Sheldon, Lipase-catalyzedreactions in ionic liquids, Org. Lett. 2 (2000) 4189–4191.
[96] S. Park, R.J. Kazlauskas, Improved preparation and use of room-temperature ionic liquids in lipase-catalyzed enantioand regioselectiveacylations, J. Org. Chem. 66 (2001) 8395–8401.
[98] K. Przybysz, E. Drzewinska, A. Stanislawska, A.W. Robak, Ionic liquidsand paper, Ind. Eng. Chem. Res. 44 (2005) 4599–4604.
[99] H. Pfruender, R. Jones, D.W. Botz, Water immiscible ionic liquids assolvents for whole cell biocatalysis, J. Biotechnol. 124 (2006) 182–190.
[100] S. Dai, Y.H. Ju, C.E. Barnes, Solvent extraction of strontium nitrate by acrown ether using room temperature ionic liquids, J. Chem. Soc. DaltonTrans. (1999) 1201–1202.
[101] J.H.J. Davis, Task-specific ionic liquids, Chem. Lett. 33 (2004)1072–1077.
[102] R. De Waele, L. Heerman, W. D’Olieslager, Electrochemistry ofuranium(IV) in acidic AlCl3 + N-(n-butyl)pyridinium chloride room-temperature molten salts, J. Electroanal. Chem. 142 (1982) 137–146.
[103] L. Heerman, R. De Waele, W. D’Olieslager, Electrochemistry andspectroscopy of uranium in basic AlCl3 + N-(n-butyl) pyridinium chlo-ride room temperature molten salts, J. Electroanal. Chem. 193 (1985)289–294.
[104] P.B. Hitchcock, T.J. Mohammed, K.R. Seddon, J.A. Zora, C.L. Hussey,E.H. Ward, 1-Methyl-3-ethylimidazolium hexachlorouranate(IV) and1-methyl-3-ethylimidazolium tetrachlorodioxo-uranate(VI): synthesis,structure, and electrochemistry in a room temperature ionic liquid, Inorg.Chim. Acta 113 (1986) L25–L26.
[105] S. Dai, Y.S. Shin, L.M. Toth, C.E. Barnes, Comparative UVVis studies ofuranyl chloride complex in two basic ambienttemperature melt systems:the observation of spectral and thermodynamic variations induced viahydrogen bonding, Inorg. Chem. 36 (1997) 4900–4902.
[106] G.M.N. Baston, A.E. Bradley, T. Gorman, I. Hamblett, C. Hardacre, J.E.Hatter, M.J.F. Healy, B. Hodgson, R. Lewin, K.V. Lovell, G.W.A. Newton,M. Nieuwenhuyzen, W.R. Pitner, D.W. Rooney, D. Sanders, K.R. Sed-don, H.E. Simms, R.C. Thied, Ionic Liquids for the Nuclear Industry: ARadiochemical, Structural and Electrochemical Investigation, Ionic Liq-uids Industrial Applications for Green Chemistry, American ChemicalSociety, Washington, DC, 2002, pp. 162–177.
[107] K. Nakashima, F. Kubota, T. Maruyama, M. Goto, Feasibility of ionicliquids as alternative separation media for industrial solvent extractionprocesses, Ind. Eng. Chem. Res. 44 (2005) 4368–4372.
[108] A.E. Visser, R.P. Swatloski, W.M. Reischert, S.T. Griffin, R.D. Rogers,Traditional extractants in nontraditional solvents: groups 1 and 2 extrac-
l Fluids 43 (2007) 150–180
tion by crown ethers in room temperature ionic liquids, Ind. Eng. Chem.Res. 39 (2000) 3596–3604.
[109] S. Chun, S.V. Dzyuba, R.A. Bartsch, Influence of structural variationin room temperature ionic liquids on the selectivity and efficiency ofcompetitive alkali metal salt extraction by a crown ether, Anal. Chem. 73(2001) 3737–3741.
[110] R.A. Bartsch, S. Chun, S.V. Dzyuba, Ionic Liquids as Novel Diluentsfor Solvent Extraction of Metal Salts by Crown Ethers, Ionic LiquidsIndustrial Applications for Green Chemistry, American Chemical Society,Washington, DC, 2002, pp. 58–68.
[111] G.-T. Wei, Z. Yang, C.J. Chen, Room temperature ionic liquid as a novelmedium for liquid/liquid extraction of metal ions, Anal. Chim. Acta 488(2003) 183–192.
[112] H. Luo, S. Dai, P.V. Bonnesen, A.C.I. Buchanan, J.D. Holbrey, N.J.Bridges, R.D. Rogers, Extraction of cesium ions from aqueous solutionsusing calix[4]arene-bis(tert-octylbenzocrown-6) in ionic liquids, Anal.Chem. 76 (2004) 3078–3083.
[113] H. Luo, S. Dai, P.V. Bonnesen, Solvent extraction of Sr2+ and Cs+ basedon room temperature ionic liquids containing monoaza-substituted crownethers, Anal. Chem. 76 (2004) 2773–2779.
[114] A.E. Visser, R.P. Swatloski, W.M. Reischert, R. Mayton, S. Sheff, A.Wierzbicki, J.H. Davis, R.D. Rogers, Task-specific ionic liquids for theextraction of metal ions from aqueous solutions, Chem. Commun. (2001)135–136.
[115] M. Wagner, M. Uerdingen, in: B. Cornils, W.A. Herrmann, I.T. Horvath,W. Leitner, S. Mecking, H. Olivier-Bourbigou, D. Vogt (Eds.), MultiphaseHomogeneous Catalysis, Wiley-VCH, Weinheim, 2005.
[116] P. Wassersheid, M. Haumann, in: D.J. Cole-Hamilton, R.P. Tooze (Eds.),Catalyst Separation, Recovery and Recycling: Chemistry and ProcessDesign, Springer, Dordrecht, 2006.
[117] K.R. Harris, M. Kanakubo, L.A. Woolf, Temperature and pressure depen-dence of the viscosity of the ionic liquids: 1-methyl-3-octylimidazoliumhexafluorophosphate and 1-methyl-3-octylimidazolium tetrafluorobo-rate, J. Chem. Eng. Data 51 (2006) 1161–1167.
[118] L.A. Blanchard, J.F. Brennecke, Recovery of organic products from ionicliquids using supercritical carbon dioxide, Ind. Eng. Chem. Res. 40 (2001)287–292.
[119] A.P.-S. Kamps, D. Tuma, J. Xia, G. Maurer, Solubility of CO2 in the ionicliquid [bmim][PF6], J. Chem. Eng. Data 48 (2003) 746–749.
[120] Z.M. Liu, W.Z. Wu, B.X. Han, Z.X. Dong, G.Y. Zhao, J.Q. Wang, T. Jiang,G.Y. Yang, Study on the phase behaviors, viscosities, and thermodynamicproperties of CO2 [C4mim][PF6]methanol system at elevated pressures,Chem. Eur. J. 9 (2003) 3897–3903.
[121] S.N.V.K. Aki, B.R. Mellein, E.M. Saurer, J.F. Brennecke, High-pressurephase behavior of carbon dioxide with imidazolium-based ionic liquids,J. Phys. Chem. B 108 (2004) 20355–20365.
[122] C. Cadena, J.L. Anthony, J.K. Shah, T.I. Morrow, J.F. Brennecke, E.J.Maginn, Why is CO2 so soluble in imidazolium-based ionic liquids? J.Am. Chem. Soc. 126 (2004) 5300–5308.
[123] Y.S. Kim, W.Y. Choi, J.H. Jang, K.-P. Yoo, C.S. Lee, Solubility mea-surement and prediction of carbon dioxide in ionic liquids, Fluid PhaseEquilib. 228–229 (2005) 439–445.
[124] M.B. Shiflett, A. Yokozeki, Solubilities and diffusivities of carbon dioxidein ionic liquids:[bmim][PF6] and [bmim][BF4], J. Am. Chem. Soc. 44(2005) 4453–4464.
[125] L.A. Blanchard, D. Hancu, E.J. Beckman, J.F. Brennecke, Green process-ing using ionic liquid and carbon dioxide, Nature 399 (1999) 28–29.
[126] M.B. King, A. Mubarak, J.D. Kin, T.R. Bott, The mutual solubilities ofwater with supercritical and liquid carbon dioxides, J. Supercrit. Fluids 5(1992) 296–302.
[127] K.L. Toews, R.M. Scholl, N.G. Smart, pH-defining equilibrium betweenwater and supercritical CO2: influence on SFE of organics and metalchelates, Anal. Chem. 67 (1995) 4040–4043.
[128] S.R. Rubero, S. Baldelli, Influence of water on the surface of hydrophilicand hydrophobic room temperature ILs, J. Am. Chem. Soc. 126 (2004)11788–11789.
[129] A. Shariati, C.J. Peters, High pressure phase behavior of systemswith ionic liquids: II. The binary system carbon dioxide + 1-ethyl-3-
ritica
S. Keskin et al. / J. of Superc
methylimidazolium hexafluorophosphate, J. Supercrit. Fluids 29 (2004)43–48.
[130] A. Shariati, C.J. Peters, High pressure phase behavior of systems withionic liquids. Part III. The binary system carbon dioxide + 1-hexyl-3-methylimidazolium hexafluorophosphate, J. Supercrit. Fluids 30 (2004)139–144.
[131] P. Husson-Borg, V. Majer, M.F.C. Gomes, Solubilities of oxygen and car-bon dioxide in butyl methyl imidazolium tetrafluoroborate as a functionof temperature and at pressures close to atmospheric pressure, J. Chem.Eng. Data 48 (2003) 480–485.
[132] M.C. Kroon, A. Shariati, M. Costantini, J. van Spronsen, G.-J. Witkamp,R.A. Sheldon, C.J. Peters, High-pressure phase behavior of systemswith ionic liquids. Part V. The binary system carbon dioxide + 1-butyl-3-methylimidazolium tetrafluoroborate, J. Chem. Eng. Data 50 (2005)173–176.
[133] M. Constantini, V.A. Toussaint, A. Shariati, C.J. Peters, I. Kikic, High-pressure phase behavior of systems with ionic liquids. Part IV. Binarysystem carbon dioxide + 1-hexyl-3-methylimidazolium tetrafluoroborate,J. Chem. Eng. Data 50 (2005) 52–55.
[134] S.G. Kazarian, N. Sakellarios, C.M. Gordon, High pressure CO2-inducedreduction of the melting temperature of ionic liquids, Chem. Commun.(2002) 1314–1315.
[135] P. Diep, K.D. Jordan, J.K. Johnson, E.J. Beekman, CO2–fluorocarbon andCO2–hydrocarbon interactions from first-principles calculations, J. Phys.Chem. A 102 (1998) 2231–2236.
[136] M.F.C. Gomes, A.A.H. Padua, Interactions of carbon dioxide with liquidfluorocarbons, J. Phys. Chem. B 107 (2003) 14020–14024.
[137] C.R. Yonker, B.J. Palmer, Investigation of CO2/fluorine interactionsthrough the intermolecular effects on the 1H and 19F shielding of CH3Fand CHF3 at various temperatures and pressures, J. Phys. Chem. A 105(2001) 308–314.
[138] K.R. Seddon, A. Stark, M.J. Torres, in: M.A. Abraham, L. Moens (Eds.),Clean Solvents, American Chemical Society, Washington, DC, ACSSymp. Ser. 819 (2002) p. 34–49.
[139] K.I. Gutkowski, A. Shariati, C.J. Peters, High-pressure phase behav-ior of the binary ionic liquid system 1-octyl-3-methylimidazoliumtetrafluoroborate + carbon dioxide, J. Supercrit. Fluids 39 (2006)187–191.
[140] M.C. Kroon, E.K. Karakatsani, I.G. Economou, G.-J. Witkamp, C.J.Peters, Modeling of the carbon dioxide solubility in imidazolium-basedionic liquids with the tPC-PSAFT equation of state, J. Phys. Chem. B 110(2006) 9262–9269.
[141] J.C. De la Fuente Badilla, C. Peters, J. de Swaan Arons, Volume expansionin relation to the gas–antisolvent process, J. Supercrit. Fluids 17 (2000)13–23.
[142] A. Shariati, C.J. Peters, High-pressure phase behavior of systems withionic liquids: measurements and modeling of the binary system fluoro-form + 1-ethyl-3-methylimidazolium hexafluorophosphate, J. Supercrit.Fluids 25 (2003) 109–117.
[143] W. Wu, J. Zhang, B. Han, J. Chen, Z. Liu, T. Jiang, J. He, W. Li, Solubilityof room-temperature ionic liquid in supercritical CO2 with and withoutorganic compounds, Chem. Commun. (2003) 1412–1413.
[144] W. Wu, W. Li, B. Han, T. Jiang, D. Shen, Z. Zhang, D. Sun, B. Wang,Effect of organic cosolvents on the solubility of ionic liquids in scCO2,J. Chem. Eng. Data 49 (2004) 1597–1601.
[145] Z.F. Zhang, W.Z. Wu, H.X. Gao, B.X. Han, B. Wang, Y. Huang, Triphasebehavior of ionic liquid–water–CO2 system at elevated pressures, Phys.Chem. Chem. Phys. 6 (2004) 5051–5055.
[146] Z.F. Zhang, W.Z. Wu, B. Wang, J. Chen, D. Shen, B. Han, High-pressurephase behavior of CO2/acetone/ionic liquid system, J. Supercrit. Fluids40 (2007) 1–6.
[147] S.G. Kazarian, B.J. Briscoe, T. Welton, Combining ionic liquids andsupercritical fluids: in situ ATR–IR study of CO2 dissolved in two ionic
liquids at high pressures, Chem. Commun. (2000) 2047–2048.
[148] Z. Gu, L.A. Blanchard, D. Hancu, E.J. Beckman, J.F. Brennecke,Environmentally-benign ionic liquid/CO2 biphasic system for separa-tions and reactions, in: Proceedings of 5th International Symposium onSupercritical Fluids, Atlanta, USA, 2000.
l Fluids 43 (2007) 150–180 179
[149] T. Al-Sahhof, R.S. Al-Ameeri, S.E. Hamam, Bubble point measurementsfor the ternary system: CO2, cyclohexane and naphthalene, Fluid PhaseEquilib. 53 (1989) 31–37.
[150] W.L. Weng, M.J. Lee, Vapor–liquid equilibrium for binary systems con-taining a heavy liquid and a dense fluid, Ind. Eng. Chem. Res. 31 (1992)2769–2773.
[151] A.M. Scurto, S.N.V.K. Aki, J.F. Brennecke, CO2 as a separation switchfor ionic liquid/organic mixtures, J. Am. Chem. Soc. 124 (2002)10276–10277.
[152] A.M. Scurto, S.N.V.K. Aki, J.F. Brennecke, Carbon dioxide induced sep-aration of ionic liquids and water, Chem. Commun. (2003) 572–573.
[154] V. Najdanovic-Visak, L.P.N. Rebelo, M.N. Da Ponte, Liquid–liquidbehavior of ionic liquid–1-butanol–water and high pressure CO2-inducedphase changes, Green Chem. 7 (2005) 443–450.
[155] S.V. Dzyuba, R.A. Bartsch, Recent advances in applications of room-temperature ionic liquid/supercritical CO2 systems, Angew. Chem. Int.Ed. 42 (2003) 148–150.
[156] M.F. Sellin, P.B. Webb, D.J. Cole-Hamilton, Continuous flow homoge-neous catalysis: hydroformylation of alkenes in supercritical fluid–ionicliquid biphasic mixtures, Chem. Commun. (2001) 781–782.
[157] P.B. Webb, M.F. Sellin, T.E. Kunene, S. Williamson, A.M.Z. Slawin, D.J.Cole-Hamilton, Continuous flow hydroformylation of alkenes in super-critical fluid–ionic liquid biphasic systems, J. Am. Chem. Soc. 125 (2003)15577–15588.
[158] U. Hintermair, G. Zhao, C.C. Santini, M.J. Muldoona, D.J. Cole-Hamilton, Supported ionic liquid phase catalysis with supercritical flow,Chem. Commun. (2007) 1462–1464.
[159] R.A. Brown, P. Pollet, E. McKoon, C.A. Eckert, C.L. Liotta, P.G. Jessop,Asymmetric hydrogenation and catalyst recycling using ionic liquid andsupercritical carbon dioxide, J. Am. Chem. Soc. 123 (2001) 1254–1255.
[160] P.G. Jessop, R.R. Stanley, R.A. Brown, C.A. Eckert, C.L. Liotta, T.T. Ngo,P. Pollet, Neoteric solvents for asymmetric hydrogenation: supercriticalfluids, ionic liquids, and expanded ionic liquids, Green Chem. 5 (2003)123–128.
[161] F. Liu, M.B. Abrams, R.T. Baker, W. Tumas, Phase-separable catalysisusing room temperature ionic liquids and supercritical carbon dioxide,Chem. Commun. (2001) 433–434.
[162] A. Bosmann, G. Francio, E. Janssen, M. Solinas, W. Leitner, P. Wasser-scheid, Activation, tuning, and immobilization of homogeneous catalystsin an ionic liquid/compressed CO2 continuous-flow system, Angew.Chem. Int. Ed. 40 (2001) 2697–2699.
[163] D. Ballivet-Tkatchenko, M. Picquet, M. Solinas, G. Francio, P. Wasser-scheid, W. Leitner, Acrylate dimerisation under ionic liquid–supercriticalcarbon dioxide conditions, Green Chem. 5 (2003) 232–235.
[164] Z. Hou, B. Han, L. Gao, T. Jiang, Z. Liu, Y. Chang, X. Zhang, J.He, Wacker oxidation of 1-hexene in 1-n-butyl-3-methylimidazoliumhexafluorophosphate ([bmim][PF6]), supercritical (SC) CO2, and SCCO2/[bmim][PF6] mixed solvent, New J. Chem. 26 (2002) 1246–1248.
[165] H. Kawanami, A. Sasaki, K. Matsui, Y. Ikushima, A rapid and effectivesynthesis of propylene carbonate using a supercritical CO2–ionic liquidsystem, Chem. Commun. (2003) 896–897.
[166] L. Gao, T. Jiang, G. Zhao, T. Mu, W. Wu, Z. Hou, B. Han, Transesteri-fication between isoamyl acetate and ethanol in supercritical CO2, ionicliquid, and their mixture, J. Supercrit. Fluids 29 (2004) 107–111.
[167] G. Zhao, T. Jiang, W. Wu, B. Han, Z. Liu, H. Gao, Electro-oxidation ofbenzyl alcohol in a biphasic system consisting of supercritical CO2 andionic liquids, J. Phys. Chem. B 108 (2004) 13052–13057.
[168] B. Yoon, C.H. Yen, S. Mekki, S. Wherland, C.M. Wai, Effect of wateron the heck reactions catalyzed by recyclable palladium chloride in ionicliquids coupled with supercritical CO2 extraction, Ind. Eng. Chem. Res.
45 (2006) 4433–4435.
[169] D.J. Cole-Hamilton, T.E. Kunene, P.B. Webb, in: B. Cornils, W.A. Her-rmann, I.T. Horvath, W. Leitner, S. Mecking, H. Olivier-Bourbigou, D.Vogt (Eds.), Multiphase Homogeneous Catalysis, Wiley-VCH, Wein-heim, 2005.
1 ritica
80 S. Keskin et al. / J. of Superc
[170] C.M. Gordon, W. Leitner, in: D.J. Cole-Hamilton, R.P. Tooze (Eds.),Catalyst Separation, Recovery and Recycling: Chemistry and ProcessDesign, Springer, Dordrecht, 2006.
[171] J.-Q. Wang, D.-L. Kong, J.-Y. Chen, F. Cai, L.-N. He, Synthesis of cycliccarbonates from epoxides and carbon dioxide over silica-supported qua-ternary ammonium salts under supercritical conditions, J. Mol. Catal. A:Chem. 249 (2006) 143–148.
[172] R. Ciriminna, P. Hesemann, J.J.E. Moreau, M. Carraro, S. Campestrini,M. Pagliaro, Aerobic oxidation of alcohols in carbon dioxide with silica-supported ionic liquids doped with perruthenate, Chem. Eur. J. 12 (2006)5220–5224.
[173] P. Lozano, T. de Diego, D. Carrie, M. Vaultier, J.L. Iborra, Continuousgreen biocatalytic processes using ionic liquids and supercritical carbondioxide, Chem. Commun. (2002) 692–693.
[174] P. Lozano, T. de Diego, D. Carrie, M. Vaultier, J.L. Iborra, Lipase catalysis
in ionic liquids and supercritical carbon dioxide at 150 ◦C, Biotechnol.Prog. 19 (2003) 380–382.
[175] M.T. Reetz, W. Wiesenhofer, G. Francio, W. Leitner, Biocatalysis inionic liquids: batchwise and continuous flow processes using supercriticalcarbon dioxide as the mobile phase, Chem. Commun. (2002) 992–993.
l Fluids 43 (2007) 150–180
[176] P. Lozano, T. de Diego, S. Gmouh, M. Vaultier, J.L. Iborra, Synthesisof glycidyl esters catalyzed by lipases in ionic liquids and supercriticalcarbon dioxide, J. Mol. Catal. A: Chem. 214 (2004) 113–119.
[177] P. Lozano, T. de Diego, D. Carrie, M. Vaultier, J.L. Iborra, Criteria todesign green enzymatic processes in ionic liquid/supercritical carbondioxide systems, Biotechnol. Prog. 20 (2004) 661–669.
[178] F.J. Hernandez, A.P. de los Rıos, D. Gomez, M. Rubio, G. Vıllora, Anew recirculating enzymatic membrane for ester synthesis in ionic liq-uid/supercritical carbon dioxide biphasic systems, Appl. Catal. B 67(2006) 121–126.
[179] P. Lozano, T. de Diego, T. Sauer, M. Vaultier, S. Gmouh, J.L. Iborra, On theimportance of the supporting material for activity of immobilized Candidaantartica lipase B in ionic liquid/hexane and ionic liquid/supercriticalcarbon dioxide biphasic media, J. Supercrit. Fluids 40 (2007) 93–100.
[180] M.T. Reetz, W. Wiesenhofer, G. Francio, W. Leitner, Continuous flow
enzymatic kinetic resolution and enantiomer separation using ionic liq-uid/supercritical carbon dioxide media, Adv. Syn. Catal. 345 (2003)1221–1228.
[181] S. Keskin, Cleaning of contaminated soils using ILs and supercriticalcarbon dioxide, M.S. Thesis, Bogazici University, Istanbul, Turkey, 2006.
