Extraction of pentachlorophenol and dichlorodiphenyltrichloroethane from aqueous solutions using ionic liquids

Extraction of pentachlorophenol and dichlorodiphenyltrichloroethane from aqueous solutions using ionic liquids

Journal of Industrial and Engineering Chemistry 18 (2012) 1983–1996 Contents lists available at SciVerse ScienceDirect Journal of Industrial and Eng...

1MB Sizes 0 Downloads 51 Views

Journal of Industrial and Engineering Chemistry 18 (2012) 1983–1996

Contents lists available at SciVerse ScienceDirect

Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec

Extraction of pentachlorophenol and dichlorodiphenyltrichloroethane from aqueous solutions using ionic liquids Santhi Raju Pilli, Tamal Banerjee, Kaustubha Mohanty * Department of Chemical Engineering, Indian Institute of Technology Guwahati, Guwahati – 781039, Assam, India

A R T I C L E I N F O

Article history: Received 9 March 2012 Accepted 23 May 2012 Available online 31 May 2012 Keywords: COSMO-RS Endocrine disruption compounds Liquid–liquid extraction Gaussian 03 Ionic liquids Wastewater treatment

A B S T R A C T

This work reports the judicious screening of various ionic liquids for the extraction of pentachlorophenol (PCP) and dichlorodiphenyltrichloroethane (DDT) from aqueous solutions. COSMO-RS model was used for the prediction of selectivity of these compounds in aqueous medium at infinite dilution. Initial benchmarking studies were done by comparing experimental logarithmic octanol–water partition coefficient values (log KOW) with COSMO-RS predicted data on 13 poly chlorinated biphenyls (PCB’s) similar in structure to that of PCP and DDT. 1015 possible ILs were then used to determine the best extractant for the removal of these two EDCs from aqueous solution. ß 2012 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.

1. Introduction In recent years, the contamination of the environment has lead to discomfort, instability, disorder leaving harmful impact on physical system, and living organism. Pentachlorophenol (PCP) and dichlorodiphenyltrichloroethane (DDT) are some of the harmful environmental chemicals due to their toxicity and pollution with dioxins. PCP is found in two forms as a sodium salt of PCP and is produced by the chlorination of phenol in the presence of a catalyst. It dissolves in water easily. DDT produced by the reaction of CCl3CHO (chloral) with chlorobenzene (C6H5Cl) in presence of sulfuric acid acting as a catalyst is almost insoluble in water but has a superior solubility in most of the organic solvents, oils and fats. Further both PCP and DDT are the commercially well-known synthetic pesticides or insecticides [1–3]. Apart from their advantages in daily life, there are health effects because of its endocrine disruption nature. Various endocrine disrupting compounds along with their effect and potential source have been reported in recent years [4] and are presented in Table 1 [5]. Thus, from a physiological point of view, an endocrine disrupting substance is a compound also called environmental hormone, either natural or synthetic, which, through environmental or inappropriate developmental exposures, alters the hormonal and homeostatic systems, thereby responding to its environment.

* Corresponding author. Tel.: +91 361 2582267; fax: +91 361 2582291. E-mail address: [email protected] (K. Mohanty).

Some EDCs/pharmaceuticals and personal care products are more polar than current regulated polyaromatic contaminants. Health effects associated with EDCs includes a range of reproductive problems (declining sperm counts), reduced fertility, male and female reproductive tract abnormalities, and skewed male/female sex ratios, loss of fetus, menstrual problems [1], changes in hormone levels; early puberty; brain and behavior problems; impaired immune functions; cryptorchism and hypospadias [6], breast cancers [7], and abnormal developments in wildlife species. For centuries, farmers have found procreative problems in female sheep and cows grazing on pastures rich in certain clover species, which later were found to contain estrogenic compounds such as coumestrol [8]. As water and wastewater are some of the main routes of exposure to EDCs, it is essential to determine at what levels EDCs are found in these media and how these levels may be reduced. Drinking water treatment mainly relies upon adsorptive [9] and oxidative [10] processes to remove or change from one form or medium into another. Several physico-chemical methods such as chemical precipitation, lime coagulation, ion exchange, reverse osmosis, biosorption and adsorption etc. [11] have been advised for selected groups of EDCs/pharmaceuticals and personal care products (PPCPs), heavy metals, pesticides, and herbicides suggesting coagulation, sedimentation, and filtration for minimal levels of removal [12,13]. Other processes include, adsorption [14], ultrasound, pyrolysis, membrane based techniques [15] etc., which removes EDCs from aqueous solution. However, these conventional physico-chemical remediation methods suffer from various disadvantages. However, most of these processes have some

1226-086X/$ – see front matter ß 2012 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jiec.2012.05.017

1984

S.R. Pilli et al. / Journal of Industrial and Engineering Chemistry 18 (2012) 1983–1996

Nomenclature

List of symbols solvent selectivity S solvent selectivity at infinite dilution S1 ij C solvent capacity effective contact surface area of a segment in A˚2 aeff OR Ci concentration of component i in octanol rich phase concentration of component i in water rich phase CiWR OR concentration of component i in octanol rich phase Ctot WR total concentration of component in water rich Ctot phase OR Ctot total concentration of component in octonol rich phase mole fraction of i in octanol rich phase xOR i mole fraction of i in water rich phase xWR i molecular surface on atom a aa chb hydrogen bonding interaction constant misfit interaction energy Emisfit hydrogen bonding interaction energy Ehb van der Waal (vdW) interaction energy EvdW Boltzmann constant K octanol–water partition coefficient KOW molecular weight M probabilistic surface charge distribution for pure p i (s ) component i normalized area parameter q ri normalized volume parameter mole fraction in liquid phase xi temperature in K T coordination number = 10 Z R universal gas constant (kcal/(mol K)) room-temperature ionic liquid RTIL ionic liquid IL infinite dilution activity coefficient IDAC supported ionic liquid membrane SLIM screening charge densities (e/A˚2) SCD pentachlorophenol PCP dichlorodiphenyltrichloroethane DDT polychlorinated biphenyl PCB endocrine disrupting chemical EDC liquid liquid extraction LLE permissible exposure limits PEL Greek symbols s screening charge density in e/A˚2 shb cut-off screening charge density for hydrogen bonding in e/A˚2

Gi(s) Gs(s) gi/s g i1 gi go g1 i;OR

g1 i;WR a0

segment activity coefficient for pure component i segment activity coefficient for mixture (s) component activity coefficient in the mixture (s) infinite dilution activity coefficient of component i chemical potential of component i in mixture chemical potential of pure component activity coefficient of component i in octanol at infinite dilution activity coefficient of component i in water at infinite dilution misfit constant (kcal A˚4/mol e2)

limitations such as, incomplete removal, cost effective, generation of sludge, more time and high energy requirements [11]. Bioremediation process which is one of the widely used process, involves the action of live and active microorganisms to degrade or remove the organic pollutant by transmuting them into a toxic substances, specifically, carbon dioxide and water [16]. Though the process is eco-friendly, low-cost and highly competent it cannot process the pollutants which are not readily biodegradable. Some of the EDCs are highly resistant to bioremediation [17,18]. Thus, it is essential for an alternate process which can deal with the various issues that are not addressed by other technique. Table 1 describes some common organic compounds found in the environment that has been associated with endocrine disruption. Extraction methods based on solvent are useful in the selective removal of EDCS from aqueous solution. The use of organic solvents like liquefied dimethyl ether is reported to perform exceptionally well for phenol extraction [19]. Compared to conventional steam-based distillation methods of separation, solvent-extraction processes are generally preferred. Problem associated with the earlier is formation azeotropes which is very difficult to separate in normal conditions. In addition, the volatile organic solvents (VOCs) in use for the extraction of pollutants are usually toxic and flammable. In order to get over these difficulties, there is a need to found cleaner and more effective means of achieving this separation using desirable extractants. A very good alternative for the volatile organic solvent as extractant is a room-temperature ionic liquid (RTIL) better known as ‘Green Solvent’. They are usually prepared by a combination of organic cations and inorganic or organic anions depending upon their need. Their properties can be greatly varied by appropriate combinations of different cations and anions. The cation and anion can be varied to obtain an IL with the desired chemical and physical properties. At ambient conditions, they remain as liquids with negligible vapor pressure and are non-flammable. ILs have versatile properties such as: low melting point (<100 8C), chemical and thermal stability, very low vapor pressure, non-flammable, high ionic and thermal conductivity, heat capacity, nonvolatile, noncorrosive and they possess excellent solubility for both organic and inorganic compounds. Admirable results have been published for the practical application of ILs in separation and extraction of various compounds, e.g. substituted benzene derivatives, organic acids, antibiotics, food and bio-products, poly-cyclic aromatic hydrocarbons, amino acids, paints and dyes, sulfur compounds in diesel oil, gasoline and anabolic androgenic steroids etc. A judicious selection of cation and anion combination needs to be under-taken in order to screen such a large group of ILs. This work reports the judicious screening of various ILs for the extraction of PCP and DDT from aqueous solutions. These chemicals are resistant and difficult to bio-degrade. A quantum chemical based ‘‘COnductor-like Screening MOdel for Real Solvents’’ (COSMO-RS) was used for the prediction of selectivity of these compounds in aqueous medium at infinite dilution. Computational-based screening methods like the COSMO-RS model is less time taken and more efficient when compared to other computational methods. Moreover, experimental screening takes lot of time and is very costly. In our study we adopted continuum solvation model such as COSMO [19] and its extension to real solvents COSMO-RS [20]. The selectivity at infinite dilution is important since DDT, PCP are present in trace amounts in water. This model a priori predicts the activity coefficient of a solute such as PCP or DDT in ionic liquid using the molecular structure as the only tool. A combination of 35 cations and 29 anions, which resulted in 1015 possible ILs, was used to determine the best extractant for the removal of these two EDCs from aqueous solution. Table 2 describes permissible exposure limits (PEL) of PCP and DDT [21–23].

S.R. Pilli et al. / Journal of Industrial and Engineering Chemistry 18 (2012) 1983–1996

1985

Table 1 Common organic compound found in the environment those have been associated with endocrine disruption [5]. Compound 0

0

0

2,2 ,3,4 ,5,5 -Hexachloro-4-biphenylol and other chlorinated biphenylols 40 ,7-Dihydroxy daidzein and other isoflavones, flavones, and flavonals Aldrin Alkylphenols Bisphenol A and phenolics DDE (1,1-dichoro-2,2bis(p-chlorophenyl)ethylene) DDT and metabolites Dicofol Dieldrin Diethylstilbestrol (DES) Endosulfan Hydroxy-PCB congeners Kepone (chlorodecone) Lindane (g-hexachlorocyclohexane) and other HCH isomers Lutolin, quercetin, and naringen Malathion Methoxychlor Octachlorostyrene Pentachloronitrobenzene Pentachlorophenol Phthalates and their ester compounds Polychlorinated biphenyls (PCBs) Polybrominated diphenyl ethers (PDBEs) Polycyclic aromatic hydrocarbons (PAHs) Tetrachlorodibenzo-para-dioxin and other halogenated dioxins and furans Toxaphene Tributyl tin and tin organometallic compounds Vinclozolin and metabolites Zineb Ziram

Endocrine effect

Potential source

Antiestrogenic Estrogenic

Degradation of PCBs released into the environment Natural flora

Estrogenic Estrogenic Estrogenic Antiandrogenic Estrogenic Estrogenic or antiandrogenic in top predator wildlife Estrogenic Estrogenic Estrogenic Antiestrogenic (competitive binding at estrogen receptor) Estrogenic Estrogenic and thyroid agonistic

Insecticide Industrial uses, surfactants Plastics manufacturing DDT metabolite Insecticide Insecticide Insecticide Pharmaceutical Insecticide Dielectric fluids Insecticide Miticide, insecticide

Antiestrogenic (as in uterine hyperplasia) Thryroid antagonist Estrogenic Thryroid agonist Thyroid antagonist Antiestrogenic (competitive binding at estrogen receptor) Estrogenic Estrogenic Estrogenic

Natural dietary compounds Insecticide Insecticide Electrolyte production Fungicide, herbicide Preservative Plasticizers, emulsifiers Dielectric fluid Fire retardants, including in utero exposures Combustion by-products Combustion and manufacturing (such as halogenation) by-product Animal pesticide dip Paints and coatings Fungicide Fungicide, insecticide Fungicide, insecticide

Antiandrogenic (aryl hydrocarbon receptor agonist) Antiandrogenic (aryl hydrocarbon receptor agonist) Estrogenic Sexual development of gastropods and other aquatic species Antiandrogenic Thyroid antagonist Thyroid antagonist

2. Theoretical considerations The initial structure of the chosen cations and anions along with EDC compounds DDT and PCP were drawn using visualization freeware MOLDEN [24]. GAUSSIAN03 package was used to generate the COSMO file [25]. The structures drawn in MOLDEN are used by the GAUSSIAN03 to conduct the geometry optimization. The idea is to get the true minimum energy geometry of the selected molecule. By using the geometry and frequency optimization, the COSMO file generation was obtained from the optimized structure. The Hartree–Fock [26] level of theory with 6-31G* basis set was used for initial geometry optimization on isolated cations and anions. To detect the presence of any negative or imaginary frequencies, a frequency optimization was done using the freq keyword in GAUSSIAN03 [25]. The COSMO file was generated using Gaussian 03 on the final frequency and geometryoptimized molecule via PBV 86 density functional theory. The

Table 2 Permissible exposure limits of PCP and DDT. NIOSH (PELs) guidelines Time weighted average (TWA) also expressed as TLV Immediately dangerous to life and health (IDLH) OSHA (PELs) regulations Time weighted average (TWA) Immediately dangerous to life and health (IDLH) WHO guidelines for drinking water Drinking water quality

PCP [21,22] 3

DDT [23]

0.5 mg/m

0.5 mg/m3

2.5 mg/m3

500 mg/m3

0.5 mg/m3 vacated: 0.5 mg/m3

1 mg/m3 (skin) No data

109 mg/L

1 mg/L

triple zeta valence potential (TZVP) [27] and the density fitting basis set DGA1 [28] has been used. For the prediction, a complete dissociation of ionic liquids into cations and anions was assumed elsewhere [29]. For this reason; the sigma profile i.e. the distribution of screening charge densities is simply the linear addition of the sigma profiles of the cation and anion. (Eq. (1)) pIL ðs Þ ¼ pcation ðs Þ þ panion ðs Þ

(1)

where pcation(s) and panion(s) are the sigma profiles for cation and anion respectively. The sigma profile of cation and anion are obtained separately and then added via simple algebraic addition followed by normalization. Thus this sigma profile will behave as a profile of a single molecule. This is equivalent in calculating the sigma profile of a mixture of both cation and anion. The COSMO volume and COSMO area also get added linearly along with the sigma profile. COSMO volume cavity and COSMO area cavity were found to increase with carbon chain length of all the cations (Tables 3 and 4). COSMO-RS assumes the interaction energy between two segments to be composed of two terms, misfit energy and the hydrogen bonding energy. The physical parameters such as aeff = 6.25 A˚2 (surface area of a standard segment), a0 = 8896 (kcal A˚4)/(mol e2) (misfit constant for misfit energy interaction), chb = 54,874 (kcal A˚4)/(mol e2) (constant for hydrogen-bonding interaction), and shb = 0.0085 e/A˚2 (the cut off value for hydrogen-bonding interactions) were taken from literature [30]. Hydrogen bonding is always assumed to be within two segments such that the hydrogen bond acceptor type has shb > 0.0085 e/A˚2and the hydrogen bond donor type has

S.R. Pilli et al. / Journal of Industrial and Engineering Chemistry 18 (2012) 1983–1996

1986 Table 3 List of cations studied in this work. S. no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

Name of cation 1-Octylquinolinium 1-Ethtyl-3-methylimidazolium 1-Butyl-3-methylimidazolium 1-Methyl-3-pentylimidazolium 1-Hexyl-3-methylimidazolium 1-Octyl-3-methylimidazolium 1,3-Dimethylimidazolium Methyl-methyl-methyl-imidazolium 1-Ethyl-2,3-dimethylimidazolium 1-Butyl-2,3-dimethylimidazolium 1-Ethyl-4-methylpyridinium 1-Butyl-4-methylpyridinium 1-Hexyl-4-methylpyridinium 1-Octyl-4-methylpyridinium 1,1-Dimethyl-pyrrolidinium 1-1-Dipropyl-pyrrolidinium 1-Butyl-1-ethyl-pyrrolidinium 1-Butyl-1-methyl-pyrrolidinium 1-Hexyl-1-methyl-pyrrolidinium 1-Octyl-1-methyl-pyrrolidinium Benzyltrimethyl-ammonium Butyltrimethyl-ammonium Dimethyl-ammonium 2-Hydroxyethyl-trimethylammonium Methyl-ammonium Ethyl-ammonium Propyll-ammonium Hexyldiethylmethylphosphonium Hexyltrimethylphosphonium Octyltrimethylphosphonium Octyldiethylmethylphosphonium Diethyldimethylpropylphosphonium Tributylmethylphosphonium Tetrabutylphosphonium Triisobutylmethylphosphonium

Acronym +

[C8CHIN] [C2MIM]+ [C4MIM]+ [C5MIM]+ [C6MIM]+ [C8MIM]+ [C1MIM]+ [C1DMIM]+ [C2DMIM]+ [C4DMIM]+ [C2MPY]+ [C4MPY]+ [C6MPY]+ [C8MPY]+ [DMPYR]+ [DPPYR]+ [BEPYR]+ [BMPYR]+ [HMPYR]+ [OMPYR]+ [BETNH]+ [BTNH]+ [DMNH]+ [HeTNH]+ [MNH]+ [ENH]+ [PNH]+ [HDEMP]+ [HTMP]+ [OTMP]+ [ODEMP]+ [DEDMPP]+ [TBMP]+ [TBP]+ [TIBMP]+

Chemical formulae/mol. wt.

COSMO area (A˚2)

COSMO volume (A˚3)

C51H61ClNO2/897.3 C6H11N2+/111.164 C8H17N2+/139.218 C9H17N2+/153.244 C10H19N2+/167.271 C12H23N2+/195.186 C5H9N2+/97.138 C6H11N2+/111.164 C7H13N2+/125.191 C9H17N2+/153.244 C8H12N2+/122.187 C10H16N+/150.24 C12H20N+/178.293 C14H24N+/206.347 C6H14N+/100.182 C10H22N+/156.288 C10H22N+/156.288 C9H20N+/142.261 C11H24N+/170.314 C13H28N+/198.368 C10H10N+/150.24 C7H18N+/116.143 C2H8N+/46.091 C5H14NO+/104.17 CH6N+/32.064 C2H8N+/46.091 C3H10N+/60.118 C11H26P+/189.3 C9H22P+/161.24 C12H22P+/204.33 C13H30P+/217.35 C9H22P+/161.24 C13H30P+/217.35 C16H36P+/259.43 C13H30P+/217.35

1150.38 599.80 746.01 821.23 892.55 1038.47 525.12 517.23 652.90 802.99 649.44 799.65 951.50 1102.38 554.00 821.65 822.70 723.57 858.05 1025.58 758.61 692.16 348.10 581.00 268.88 345.02 420.11 1050.89 914.95 1061.34 1197.27 876.01 1195.22 1413.94 1138.84

2165.47 1006.59 1279.15 1414.94 1549.42 1822.07 871.17 860.52 1131.94 1403.62 1123.80 1395.80 1669.13 1941.82 948.93 1492.03 1491.18 1299.08 1547.25 1817.48 1380.66 1182.72 511.33 974.99 366.56 503.84 640.29 1831.43 1561.10 1832.54 2103.81 1553.30 2102.03 2514.52 2087.54

Table 4 List of anions studied in this work. S. No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

Name of anion Acetate Tetracyanoborate Tetrafluoroborate Bis(methylsulfonyl)amide Bis(oxalato (2-))borate Bis(trifluoromethylsulfonyl)amide Trifluoro-methylsulfonate Methylsulfonate Methyl sulfate Ethylsulfate Octylsulfate Chloride Dimethylphosphate Hydrogensulfate N-methylsulfonylacetamide 2-(2-methoxyethoxy)ethylsulfate Dicyanamide Hexafluorophosphate Thiocyanate p-toluenesulfonate Salicylate Nitrate Trifluoroacetate Diethyl phosphate Decanoate Dibutyl phosphate Bromide Dihydrogen phosphate Perchlorate

Acronym

Chemical formulae/mol. wt. (g/mol)

COSMO area (A˚2)

COSMO volume (A˚3)

[OAc][B(CN)4][BF4][BMA][BOB][BTA][CF3SO3][CH3SO3][CH3SO4][C2H5SO4][C8H17SO4][Cl][DMPO4][HSO4][MAcA][MDEGSO4][N(CN)2][PF6][SCN][TOS][Sal][NO3][CF3COO][DEP][DEC][DBP][Br][DHPO4][ClO4]-

C2H3O2/59.014 C4BN4/114.881 BF4/86.805 C2H6NO4S2/172.204 C8H14BO4/249.004 C2F6NO4S2/280.147 CF3O3S/148.953 CH3O3S/95.098 CH4O4S/112.105 C2H6O4S/126.132 C8H18O4S/210.291 

324.56 584.76 384.80 605.84 619.64 771.89 462.25 383.90 420.3 491.58 945.45 188.59 516.63 346.73 557.06 765.83 359.89 430.83 309.73 678.49 564.18 269.87 403.68 664.86 925.51 962.02 209.37 365.02 334.60

475.39 975.27 567.46 1036.03 1056.41 1328.92 734.16 588.65 658.09 793.51 1610.46 243.53 827.84 514.95 923.20 1340.19 545.93 674.69 465.90 1197.19 995.26 366.72 621.01 1101.36 1565.97 1644.95 284.87 539.57 485.12

Cl /35.454 C2H7O4P/12126.048 H2O4S/98.078 C3H7NO3S/137.158 C5H12O6S/200.210 C2N3/60.8 F6P/144.965 CNS/58.083 C7H7O3S/171.194 C7H5O3/137.113 NO3/62.01 C2F3O2/113.016 C4H11O4P/154.102 C10H19O2/171.257 C18H19O4P/210.208 Br/79.905 H3O4P/97.995 ClO4/99.451

S.R. Pilli et al. / Journal of Industrial and Engineering Chemistry 18 (2012) 1983–1996

shb < 0.0085 e/A˚2. The equations for misfit energy (Eq. (2)) and hydrogen bonding energy (Eq. (3)) are given below. Misfit energy: Emisfit ðs ; s 0 Þ ¼

written as: S12;max ¼ S1 12 ¼



0 2

0

aeff a ðs þ s Þ 2

(2)

3=2 aeff = 0 Þ,

1

4

where a ¼ ð0:64  0:3  e and e0 = 2.395  10 (kcal A˚) is the permittivity in free space. Hydrogen bond interaction energy:

2

(e mol)/

1987

g1 2 g1 1

IL phase 

g1 1 g1 2

water phase

 

g1 2 g1 1

IL phase

where IL stands for ionic liquid, g 1 1 is the activity coefficient of EDC species at infinite dilution and g 1 2 activity coefficient of water 1 component at infinite dilution, usually g 1 1 or g 2 is defined as: 1 1 g1 1 or g 2 ¼ lim g i

(8)

x!0

0

(7)

0

Ehb ðs ; s Þ ¼ aeff chb ðs ; s Þ ¼ aeff chb minf0; minð0; s don þ s hb Þmaxð0; s acc  s hb Þg (3) The activity coefficient of any component in the mixture is then given by [40] X lng ilS ¼ ni pi ðs Þ½lnG ðs Þ  G i ðs Þ þ lng SG (4) ilS s

lng SG ilS ¼ ln

fi xi

þ

z u fX q ln i þ li  i xl 2 i fi xi j j j

(5)

where xi q i xr z ; fi ¼ P i i and li ¼ ½ðr i  qi Þ  ðr i  lÞ 2 x q x r j j j j j j

ui ¼ P

(6)

Here, ri and qi are normalized using the volume and surface area of a functional group [31] and ln G(s) and ln Gi (s) are activity coefficients of the segment in the mixture and the compound respectively. The details of COSMO-RS methodology are given by Sandler and Lin [32] and Klamt and Schuu¨u¨rmann [20]. We have re-implemented the COSMO-SAC model as given by Sandler and Lin [32]. The applications of our model are given in our earlier work [30,33]. The prediction of activity coefficients at infinite dilution was carried since solvent effects were observed maximum when the concentration of solute (EDCs in this case) is infinitely small. However, infinite dilution also applies to the situation where the presence of a trace amount of EDCs poses the biggest problem. The selectivity (S) is defined as the ratio of the water species in IL rich phase (extract) and to the composition of EDC (effluent) in IL rich phase [34]. The operational expression mathematically can be

The capacity is directly used for evaluating the amount of ILs needed for the removal of EDC species [30]. It signifies, the ability to dissolve maximum amount of EDC species in the solvent. The capacity (C) can be expressed in terms of activity at infinite dilution: C1 12 ¼



1

IL phase

(9)

g1 1

The subscripts ‘1’ and ‘2’ indicates EDC species in IL phase and ionic liquid respectively. The performance index is defined as the product of selectivity at infinite dilution and capacity at infinite dilution for any solvent. This is represented by: PI ¼

1 C12

¼ S12;max ¼



1

IL phase

g1 1





g1 g1 2 ¼  2 2  1 g1 g1 1

!IL phase (10)

The selectivity and capacity strongly depends on solvent–solute interaction effect [35]. In general, the extraction capacity depends on anion volume and interaction energy between cation and anion. All activity coefficients calculated in the present study are at infinite dilution. The interaction becomes stronger between the solute and the ionic liquid when the value of the activity coefficient is less than unity. The selectivity values computed using COSMORS provide an efficient method for the screening of ILs. A screening of 1015 possible ionic liquids was performed to determine the best extractant. A total of 35 cations are used in combination with 29 anions. Tables 3 and 4 describe the list of cations and anions studied in this work respectively with its chemical structure, chemical formulae, molecular weight, COSMO area and COSMO volume.

Table 5 Reported properties and structures of PCP and DDT. No.

Name and chemical structure

1

Cl

Cl

HO

pKd [37,38]

Water solubility @ 25 8C [37,39,40] (mg/L) @ ionic strength (mM)

C6Cl5OH/

4.6–5.3

44 @ 5.8, 0.0003 57 @ 5.6, 0.2

266.34

4.5  0.3

C14H9Cl5/354.49

0.05–0.1

260 @ 6.6, 0.002 273 @ 6.3, 0.2 1275 @ 7.2, 0.01 [email protected] 7.2, 0.2 Highly soluble in organic solvents

log(KoW)preda

log(KoW)exp [36]

4.48

5.01

10.85

6.91

Cl

Cl

Cl

2

Cl Cl Cl

Cl Cl

a

Chemical formulae/mol. wt. (g/mol)

Predicted using COSMO–RS.

S.R. Pilli et al. / Journal of Industrial and Engineering Chemistry 18 (2012) 1983–1996

1988

3. Results and discussion 3.1. Experimental validation Before proceeding for prediction of selectivity values, we have benchmarked the COSMO-RS model with known octanol–water partition coefficients as reported in literature [36,37]. Primary benchmarking studies have been carried out for PCP and DDT by comparing our predictions with the experimental (Table 5). Theoretical partition coefficients were calculated by the following equation:

K OW;i ¼

OR 1 OR OR Ctot g i;WR g1 CiOR Ctot xi i;WR ¼ WR ¼ WR 1 ¼ 0:148 1 WR WR g i;OR Ci Ctot xi Ctot g i;OR

(11)

Table 6 Comparison of predicted and experimental octanol–water partition coefficients. Name of PCB compound

Cl

Cl

Results from the Table 5 shows that the log(KoW)pred and log(KoW)exp values of PCP very close to each other and were 4.48 and 5.01 respectively. In case of DDT similar trend was observed but log(KoW)pred value was little higher than log(KoW)exp i.e. 10.85 and 6.91 respectively (Table 5) [37–40]. Benchmarking studies have also been carried out for 13 other polychlorinated biphenyls (PCBs) by comparing our predictions with the experimental octanol–water partition coefficients as reported by Darryl and Des [41]. The PCB’s were chosen since it represents a similar structure as that of DDT and PCP. Thus it served as another evidential basis for our comparison. Table 6 describes the activity coefficient of different PCBs in water rich phase and activity coefficient of different PCBs in octanol rich phase respectively. It also describes theoretical and experimental logarithms of octanol–water partition coefficient of different

gwater

goctanol

 1  g K OW;i ¼ 0:148 gi;WR 1

6.18E + 09

0.880

1,039,093,842

9.016

5.48

3.09E + 09

0.791

578,101,617.8

8.762

5.46

8.73E + 09

1.111

1,162,554,776

9.065

5.55

6.28E + 09

0.817

1,136,755,936

9.055

5.94

6.69E + 09

0.908

1,089,939,661

9.037

6.18

5.25E + 09

0.659

1,178,949,189

9.071

6.31

i;OR

log (KOW)preda

log (KOW)exp [41]

Cl

Cl Cl

Cl

Cl

Cl

Cl

Cl

Cl

Cl

Cl

Cl

Cl

Cl

Cl

Cl

Cl

Cl Cl Cl

Cl

Cl

S.R. Pilli et al. / Journal of Industrial and Engineering Chemistry 18 (2012) 1983–1996 Table 6 (Continued ) Name of PCB compound

Cl

1989

gwater

goctanol

 1  g K OW;i ¼ 0:148 gi;WR 1 i;OR

2.02E + 09

0.322

927,307,477.5

8.967

6.21

3.69E + 10

0.807

6,758,355,011

9.829

5.37

4.84E + 10

1.061

6,752,447,711

9.829

5.82

2.93E + 11

1.047

41,424,181,697

10.617

5.76

1.51E + 13

1.203

1.85387E + 12

12.268

7.21

1.36E + 13

1.274

1.57591E + 12

12.197

7.67

1.64E + 14

2.138

1.13182E + 13

13.053

7.52

log (KOW)preda

log (KOW)exp [41]

Cl

Cl

Cl Cl

Cl

Cl

Cl

Cl

Cl Cl

Cl Cl

Cl

Cl

Cl Cl

Cl Cl Cl

Cl

Cl Cl

Cl

Cl

Cl

Cl Cl

Cl Cl

Cl

Cl

Cl

Cl

Cl Cl

Cl Cl

Cl

Cl

Cl

Cl a

Cl

Cl

Cl

Predicted using COSMO–RS.

PCBs. The predictions agree well but a higher trend was observed with the experimental findings for PCBs. Similar higher values of deviation was observed from the work of Gerber and de Rafael [42]. In their work they obtained a ln AAD (logarithmic absolute average deviation) of 81% using optimized COSMO-RS parameters for 748 aqueous system data points. In another work by Putnam et al. [43] a 44% AAD (Absolute Average Deviation) was reported for 106 aqueous systems consisting of water with alcohol, aldehyde, alkane and aromatics respectively.

Nevertheless, the benchmarking study clearly shows that similar trend was observed when compared predicted partition coefficient log (KO–W)pred with experimental partition coefficient log (KO– W)exp. After successful benchmarking, five different groups namely imidazolium, pyridinium, pyrrolidinium, ammonium and phosphonium based cations were combined with 29 different anions and were then investigated using COSMO-RS based model for screening. It should be noted that the selectivity and capacity are

S.R. Pilli et al. / Journal of Industrial and Engineering Chemistry 18 (2012) 1983–1996

functions of infinite dilution activity coefficients (IDAC) (Eqs. (7) and (9)). 3.2. Imidazolium based cations Fig. 1 shows the selectivity values of all imidazolium cations for PCP and DDT. During the screening process, [C8MIM][Sal], [C4DMIM][C8H17SO4], [C4DMIM][BTA] gave the highest selectivity of 530.5, 512.833 and 355.124 respectively for the removal of PCP (Fig. 1(A)). Favorable anions are [PF6], [BTA], [BOB] and [B(CN)4] for all cations except for the cation [C1DMIM) (where its IDAC value was <1) due to their higher (between 4 and 36) activity coefficient of water component at infinite dilution for PCP and DDT. In case of DDT (Fig. 1(B)), selectivity pattern follows: [C4DMIM][PF6] (6.768) and [C4DMIM] [BTA] (6.407). Among all the cations [C8MIM] gave high selective value (530.5) because of its higher carbon chain length. It can be seen that irrespective of pollutant, the selectivity increases with an increase in the carbon chain length on imidazolium cation. This was due to the increasing van der Waals volumes which lead to increasing entrapment of the pollutant in the cation structure. However the selectivity of DDT is low as 550 500

C2 MIM C4 MIM C5 MIM C6 MIM C8 MIM C1 MIM C1 DMIM C2 DMIM C4 DMIM

A

450

Selectivity

400 350 300 250 200

compared to PCP. It was observed that the higher activity coefficient of water component at infinite dilution for anion [PF6] with [C4DMIM] and [C2DMIM] as 181.76 and 145.952 respectively. Anions [OAc], [Cl], [DEC] and [ClO4] were not suitable for imidazolium based cation for PCP extraction. Similarly ILs [C2DMIM][DEP], [C4DMIM][DEP], [C6MIM] [DBP], [C8MIM][DBP], [C2DMIM][DBP], [C4DMIM][DBP], [C2DMIM][DMPO4] and [C2DMIM][Br] also not suitable because of their IDAC values with PCP were found to be zero. In case of DDT extraction, anion [ClO4] was not suited as its IDAC values was zero. 3.3. Pyridinium based cations Fig. 2 shows the selectivity values for the pyridinium based cation with PCP and DDT. For pyridinium based cations, [C8MPY][MAcA] (731) (Fig. 2(A)), gave higher selectivity and the second highest selectivity values obtained in this group of IL were [C8MPY][Sal] (584), [C8MPY][C8H17SO4] (582.71), [C4MPY][Sal] (584) and [C6MPY][Sal] (533) for the removal of PCP. Anions [OAc], [Cl], [DEC], [DBP] and [ClO4] were not suitable for pyridinium based cation for PCP removal. Similarly except cation [C8MPY], anion

Selectivity

1990

150 100 50

C 2 MPY C 4 MPY C 6 MPY C 8 MPY

A

[OAc][B(CN)4][BF4][BMA][BOB][BTA][CF3SO3][CH3SO3][CH3SO4][C2H5SO4][C8H17SO4][Cl][DMPO4][HSO4][MAcA][MDEGSO4][N(CN)2][PF6][SCN][TOS][Sal][NO3][CF3COO][DEP][DEC][DBP][Br][DHPO4][ClO4]-

[OAc][B(CN)4][BF4][BMA][BOB][BTA][CF3SO3][CH3SO3][CH3SO4][C2H5SO4][C8H17SO4][Cl][DMPO4][HSO4][MAcA][MDEGSO4][N(CN)2][PF6][SCN][TOS][Sal][NO3][CF3COO][DEP][DEC][DBP][Br][DHPO4][ClO4]-

0

800 750 700 650 600 550 500 450 400 350 300 250 200 150 100 50 0

Anion

10 9

C 2MPY C 4MPY C 6MPY C 8MPY

B

8 7

Selectivity

C2 DMIM C4 DMIM C5 DMIM C6 DMIM C8 DMIM C1 MIM C1 DMIM C2 DMIM C4 DMIM

B

6 5 4 3 2 1

Anion Fig. 1. Selectivity at infinite dilution and at ambient temperature (T = 25 8C) for imidazolium based IL with (A) PCP and B) DDT.

[OAc][B(CN)4][BF4][BMA][BOB][BTA][CF3SO3][CH3SO3][CH3SO4][C2H5SO4][C8H17SO4][Cl][DMPO4][HSO4][MAcA][MDEGSO4][N(CN)2][PF6][SCN][TOS][Sal][NO3][CF3COO][DEP][DEC][DBP][Br][DHPO4][ClO4]-

0 [OAc][B(CN)4][BF4][BMA][BOB][BTA][CF3SO3][CH3SO3][CH3SO4][C2H5SO4][C8H17SO4][Cl][DMPO4][HSO4][MAcA][MDEGSO4][N(CN)2][PF6][SCN][TOS][Sal][NO3][CF3COO][DEP][DEC][DBP][Br][DHPO4][ClO4]-

Selectivity

Anion 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0

Anion Fig. 2. Selectivity at infinite dilution and at ambient temperature (T = 25 8C) for pyridinium based IL with (A) PCP and B) DDT.

S.R. Pilli et al. / Journal of Industrial and Engineering Chemistry 18 (2012) 1983–1996

3.4. Pyrrolidinium based cations In the pyrrolidinium based ILs: [DPPYR][C8H17SO4] and [BEPYR][C8H17SO4] gave highest selectivity of 984.66 and 980.33 respectively for PCP as shown in Fig. 3(A). The selectivity values of DDT were very poor when compared with PCP. Among them the highest were: [BEPYR][BTA] (5.417) and [DPPYR][BTA] (4.961) as 1100 1000

DMPYR DPPYR BEPYR BMPYR HMPYR OMPYR

A

900 800

Selectivity

700 600 500

shown in Fig. 3(B). Out of imidazolium, pyridinium and pyrrolidinium cations, the pyrrolidinium based cations gave the highest selectivity. Anions [Cl], [DEC] and [ClO4] were not suitable for pyrrolidinium based cation for the extraction of PCP. Similarly cations [DPPYR] and [BEPYR] combined with anions [DMPO4], [MAcA], [Sal], [DEP], [DBP], [Br] and [DHPO4] were also undesirable for PCP extraction. Except cation [DMPYR], anion [OAc] was undesirable in this group. However as observed earlier, DDT gave poor selectivity for most cations. Except cation [DMPYR], all pyridinium based cations gave higher IDAC values for water component and the favorable anions were [PF6], [BTA], [BOB] and [B(CN)4]. 3.5. Ammonium based cations Fig. 4 shows the selectivity values of different combinations of anions and cations of ammonium group. In case of ammonium

Selectivity

[DEP] was also not suitable for the extraction of PCP. In case of DDT (Fig. 2(B)), selectivity of [C8MPY][PF6] was 9.081 followed by [C8MPY][BTA] where the selectivity value was 5.454. Other ILs with moderate selectivities in this case were [C6MPY][PF6] and [C6MPY][BTA] respectively and its selectivity values were 3.494 and 3.443. The cation [C8MPY] has higher selectivity (731) because of its higher carbon chain length. As compared to imidazolium cations, the pyridinium cations have higher selectivity for PCP. However the increasing trend with size of cation was also observed for pyridinium cations.

400 300 200 100

[OAc][B(CN)4][BF4][BMA][BOB][BTA][CF3SO3][CH3SO3][CH3SO4][C2H5SO4][C8H17SO4][Cl][DMPO4][HSO4][MAcA][MDEGSO4][N(CN)2][PF6][SCN][TOS][Sal][NO3][CF3COO][DEP][DEC][DBP][Br][DHPO4][ClO4]-

Anion

Anion

5.5 5.0 4.5

3.5 3.0

BETNH BTNH DMNH HETNH MNH ENH PNH C8CHIN

4.0 3.5

Selectivity

4.0

Selectivity

4.5

DMPYR DPPYR BEPYR BMPYR HMPYR OMPYR

B

BETNH BTNH DMNH HETNH MNH ENH PNH C8CHIN

900 850 800 750 700 650 600 550 500 450 400 350 300 250 200 150 100 50 0

[OAc][B(CN)4][BF4][BMA][BOB][BTA][CF3SO3][CH3SO3][CH3SO4][C2H5SO4][C8H17SO4][Cl][DMPO4][HSO4][MAcA][MDEGSO4][N(CN)2][PF6][SCN][TOS][Sal][NO3][CF3COO][DEP][DEC][DBP][Br][DHPO4][ClO4]-

0

6.0

1991

3.0 2.5 2.0

2.5 1.5

2.0 1.5

1.0

1.0

0.5

0.5

[OAc][B(CN)4][BF4][BMA][BOB][BTA][CF3SO3][CH3SO3][CH3SO4][C2H5SO4][C8H17SO4][Cl][DMPO4][HSO4][MAcA][MDEGSO4][N(CN)2][PF6][SCN][TOS][Sal][NO3][CF3COO][DEP][DEC][DBP][Br][DHPO4][ClO4]-

Anion Fig. 3. Selectivity at infinite dilution and at ambient temperature (T = 25 8C) for pyrrolidinium based IL with (A) PCP and B) DDT.

[OAc][B(CN)4][BF4][BMA][BOB][BTA][CF3SO3][CH3SO3][CH3SO4][C2H5SO4][C8H17SO4][Cl][DMPO4][HSO4][MAcA][MDEGSO4][N(CN)2][PF6][SCN][TOS][Sal][NO3][CF3COO][DEP][DEC][DBP][Br][DHPO4][ClO4]-

0.0

0.0

Anion Fig. 4. Selectivity at infinite dilution and at ambient temperature (T = 25 8C) for ammonium based IL with (A) PCP and B) DDT.

S.R. Pilli et al. / Journal of Industrial and Engineering Chemistry 18 (2012) 1983–1996

1992

based ILs the highest selectivity values of PCP (Fig. 4(A)) were observed as [C8CHIN][MAcA] (856) and [C8CHIN][CF3COO] (823). Similarly other than cation [C8CHIN], selectivity order followed: [BTNH][Sal] (342), [BTNH][B(CN)4] (20.556) and [BTNH][Sal] (17.604). In case of DDT (Fig. 4(B)), [C8CHIN][BTA] (4.067) gave highest selectivity followed by [C8CHIN][B(CN)4] (2.675) and [BTNH][BTA] (0.265) and [BTNH][B(CN)4] (0.122). It can be seen that the selectivity values for DDT are less than unity in most of the cases, which is undesirable. Thus ammonium based cations having selectivity of < 1 are not recommended for extraction of DDT from aqueous phase. Among imidazolium, pyridinium, pyrrolidinium and ammonium based cations, pyrrolidinium based cations gave the high selectivity followed by the ammonium based cations for extraction of PCP and for DDT, pyridinium based cations gave the high selectivity followed by the pyrrolidinium based cations. 3.6. Phosphonium based cations Fig. 5 shows the selectivity values of different combinations of anions and cations of phosphonium based group. In this case

250000

A 2100

200000

1800

150000

Selectivity

Selectivity

1500 1200 900 600

100000

300 0

50000

[TBP][PF6], gave the highest selectivity among the all ILs for the removal of PCP and having selectivity of 2,43,239 which is a very high value when compared to imidazolium, pyridinium, pyrrolidinium and ammonium based ILs and is shown in Fig. 5(A). Second high selective IL in this same group was found to be [TBP][BOB] and its selectivity was (69,252). Similarly, the favorable anions for this case were [B(CN)4], [C8H17SO4], [MDEGSO4], [N(CN)2], [TOS] and [NO3]. In case of DDT (Fig. 5(B)), order of high selective ILs were [TBP][PF6] (26.18), [TBP][BTA] (24.857), [TBP][B(CN)4] (13.888) and [ODEMP][BTA] (11.108) for the extraction of PCP. It was observed that the higher activity coefficient of water component at infinite dilution for anion [PF6] with [TBP] and [ODEMP] are 48.648 and 31.175 respectively. Among of imidazolium, pyridinium, pyrrolidinium, ammonium and phosphonium based cations, phosphonium based cations gave the highest selectivity for both PCP and DDT extraction. From the Sections 3.1–3.5, the highest selective pattern of different cations groups for the removal of PCP followed the order: phosphonium > pyrrolidinium > ammonium > pyridinium > imidazolium. Fan et al. [3] and Blanchard et al. [44] described the hydrophobic nature of ionic liquids which are suitable for liquid–liquid extraction of organic chemicals from their aqueous solution. The hydrophobic nature of any component depends on the activity coefficient of the extractant. Even though selectivity values were very high their corresponding activity coefficients at infinite dilution (IDAC) were very less and most of the cases they were less than one indicating high miscibility with water. This may cause the solubility of the ionic liquid to increase in aqueous phase. Table 7 summarizes the results obtained in Sections 3.1–3.5. The ILs having higher selectivity values were chosen and the corresponding g 1 i of IL in water is also indicated. This helps as a guide in the selection of the proper IL i.e. higher values of selectivity as well as high values of g 1 i water as the requirements. 3.7. Capacity at infinite dilution

Anion

[OAc][B(CN)4][BF4][BMA][BOB][BTA][CF3SO3][CH3SO3][CH3SO4][C2H5SO4][C8H17SO4][Cl][DMPO4][HSO4][MAcA][MDEGSO4][N(CN)2][PF6][SCN][TOS][Sal][NO3][CF3COO][DEP][DEC][DBP][Br][DHPO4][ClO4]-

0

The capacities of ILs at infinite dilution were predicted using COSMO-RS model for the extraction of EDCs from the Eq. (9). Five group of cations, imidazolium, pyridinium, pyrrolidinium, ammonium and phosphonium based were predicted using COSMO-RS

Anion

27

HDEMP HTMP OTMP ODEMP DEDMPP TBMP TBP TIBMP

B

24 21

Selectivity

18 15 12 9 6

11000 10000 9000 8000

Solvent Capacity

30

7000 6000 5000 4000 3000 2000

3

1000 0

Anion Fig. 5. Selectivity at infinite dilution and at ambient temperature, (T = 25 8C) for phosphonium based IL with (A) PCP and B) DDT.

[OAc][B(CN)4][BF4][BMA][BOB][BTA][CF3SO3][CH3SO3][CH3SO4][C2H5SO4][C8H17SO4][Cl][DMPO4][HSO4][MAcA][MDEGSO4][N(CN)2][PF6][SCN][TOS][Sal][NO3][CF3COO][DEP][DEC][DBP][Br][DHPO4][ClO4]-

[OAc][B(CN)4][BF4][BMA][BOB][BTA][CF3SO3][CH3SO3][CH3SO4][C2H5SO4][C8H17SO4][Cl][DMPO4][HSO4][MAcA][MDEGSO4][N(CN)2][PF6][SCN][TOS][Sal][NO3][CF3COO][DEP][DEC][DBP][Br][DHPO4][ClO4]-

0

Anion Fig. 6. Solvent capacity values at infinite dilution and at ambient temperature (T = 25 8C) for imidazolium based IL with PCP.

S.R. Pilli et al. / Journal of Industrial and Engineering Chemistry 18 (2012) 1983–1996

1993

Table 7 Best selective ILs and their corresponding aqueous phase activity coefficient at infinite dilution.a No.

Name of the selective IL

Chosen EDCs PCP

DDT

Selectivity Imidazolium based ILs 1 2 3 4 5 6 7 Pyridinium based ILs 1 2 3 4 5 6 7 8 Pyrrolidinium bases ILs 1 2 3 4 5 6 7 8 9 Ammonium based ILs 1 2 3 4 5 Phosphonium based ILs 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 a

g1 water

Selectivity

g1 water

[C8MIM][Sal] [C4DMIM][C8H17SO4] [C4DMIM][BTA] [C8MIM][BTA] [C8MIM] [PF6] [C4DMIM][PF6] [C4DMIM][B(CN)4]

530.5 512.833 355.124 71.871 3.044 – –

0.106 0.307 33.559 8.078 7.735 – –

– – 6.407 – – 6.768 –

– – 33.559 – – 181.762 –

[C8MPY][MAcA] [C8MPY][BTA] [C8MPY][B(CN)4] [C6MPY][PF6] [C6MPY][BTA] [C8MPY][PF6] [C8MPY][BTA] [C6MPY][B(CN)4]

731 282.816 250.356 17.961 193.438 – – –

0.073 21.918 13.995 47.121 18.686 – – –

– – 2.236 – – 9.081 5.454 –

– – 13.994 – – 66.691 21.918 –

[DPPYR][C8H17SO4] [BEPYR][C8H17SO4] [BEPYR][CF3SO3] [BEPYR][B(CN)4] [BEPYR][BTA] [DPPYR][BTA] [BEPYR][BOB] [BRPYR][B(CN)4] [DPPYR][BOB]

984.666 980.333 295.552 262.654 246.954 226.160 68.228 – –

0.295 0.294 1.123 12.239 15.780 14.474 12.547 – –

– – – – 5.417 4.961

– – – – 15.780 14.474

2.226 –

12.239 –

[BTNH][Sal] [BTNH] [B(CN)4] [BTNH] [BTA] [BETNH] [B(CN)4] [BTNH][BOB]

342 20.556 17.604 – –

0.034 2.792 3.848 – –

– 0.122 0.265 0.079 0.037

– 3.848 2.792 3.424 2.502

48.648 20.776 0.531 0.185 0.364 0.341 1.356 1.167 1.127 26.761 – – – – –

26.18 12.09 – – – – – – – 24.86 13.88 11.108 8.136 7.481 5.162

48.648 20.776 – – – – – – – 26.761 18.486 19.558 31.175 15.978 13.377

[TBP][PF6] [TBP][BOB] [TBP][BMA] [ODEMP][BF4] [TBP][C8H17SO4] [TIBMP][C8H17SO4] [ODEMP][CF3SO3] [TIBMP][CF3SO3] [TBMP][CF3SO3] [TBP][BTA] [TBP][B(CN)4] [ODEMP][BTA] [ODEMP][PF6] [HDEMP][BTA] [DEDMPP][BTA]

2,43,239 69,252 5317 1858 1821 1708.5 713.9 686.47 663.1 579.25 – – – – –

Predicted using COSMO–RS.

model. Fig. 6 shows the imidazolium based cations with 29 anion combinations for the PCP species which gave high capacity. The maximum extraction capacity was as high as 10,000 for PCP. Similar values were observed for phosphonium and pyrrolidinium based cations (figure not shown here). Cations, from phosphonium group [TBP], [TBMP], [OTMP], [TIBMP], [OTMP] and [DEDMPP], from pyrrolidinium group [DPPYR], [OMPYR] and [HMPYR] and from imidazolium [C4DMIM], [C2DMIM], [C1DMIM] and [C4DMIM] gave high capacity of 10,000. These cations possess more capacity for the extraction of six membered compounds like PCP and DDT. Fig. 7 shows the capacity of phosphonium based ILs at infinite dilution for DDT which was very poor when compared to PCP. In case of DDT, the maximum extraction capacity was 7.473 for [TBP] with anion [Cl] (Fig. 7). On the contrary the lowest value was encountered for pyridinium based cations (figure not shown here) on account of DDT.

3.8. Performance index at infinite dilution Figs. 8 and 9 shows the performance index of the PCP and DDT respectively and was predicted using COSMO-RS model for all the five group of cations. By using Eq. 10 we can calculate the performance index at infinite dilution. It can be seen that PI values of cations having higher carbon chain length gave the high performance index values for PCP (Fig. 8) and the correspondingorder followed: [TBP][PF6] (1,216,195,000) > [C8CHIN] (85,60,000) > [C8MPY] (73,10,000) > [DPPYR] (46,80,000) > [C4DMIM] (35,30,000). As observed in case of capacity the performance index also followed the similar trend. Where as in case of DDT (Fig. 9) it was observed that very small performance index values obtained with anion [BTA] i.e.: [TBP] (23.08865) > [BEPYR] (1.860058) > [C8CHIN] (1.47333889) > [C8MPY] (1.35744472) > [C4DMIM] (1.223529867). It can be seen

S.R. Pilli et al. / Journal of Industrial and Engineering Chemistry 18 (2012) 1983–1996

8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0

2.0

HDEMP HTMP OTMP ODEMP DEDMPP TBMP TBP TIBMP

1.8

Performance Index

1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

[OAc][B(CN)4][BF4][BMA][BOB][BTA][CF3SO3][CH3SO3][CH3SO4][C2H5SO4][C8H17SO4][Cl][DMPO4][HSO4][MAcA][MDEGSO4][N(CN)2][PF6][SCN][TOS][Sal][NO3][CF3COO][DEP][DEC][DBP][Br][DHPO4][ClO4]-

Anion

Anion

Fig. 7. Solvent capacity values at infinite dilution and at ambient temperature (T = 25 8C) for phosphonium based IL with DDT.

A

1.00E+009

8.00E+008

HDEMP HTMP OTMP ODEMP DEDMPP TBMP TBP TIBMP

6.00E+008

4.00E+008

24

HDEMP HTMP OTMP ODEMP DEDMPP TBMP TBP TIBMP

B

20

Performance Index

1.20E+009

Performance Index

DMPYR DPPYR BEPYR BMPYR HMPYR OMPYR

A

1.6

[OAc][B(CN)4][BF4][BMA][BOB][BTA][CF3SO3][CH3SO3][CH3SO4][C2H5SO4][C8H17SO4][Cl][DMPO4][HSO4][MAcA][MDEGSO4][N(CN)2][PF6][SCN][TOS][Sal][NO3][CF3COO][DEP][DEC][DBP][Br][DHPO4][ClO4]-

Solvent Capacity

1994

16 12 8 4

2.00E+008

[OAc][B(CN)4][BF4][BMA][BOB][BTA][CF3SO3][CH3SO3][CH3SO4][C2H5SO4][C8H17SO4][Cl][DMPO4][HSO4][MAcA][MDEGSO4][N(CN)2][PF6][SCN][TOS][Sal][NO3][CF3COO][DEP][DEC][DBP][Br][DHPO4][ClO4]-

[OAc][B(CN)4][BF4][BMA][BOB][BTA][CF3SO3][CH3SO3][CH3SO4][C2H5SO4][C8H17SO4][Cl][DMPO4][HSO4][MAcA][MDEGSO4][N(CN)2][PF6][SCN][TOS][Sal][NO3][CF3COO][DEP][DEC][DBP][Br][DHPO4][ClO4]-

0

Anion

Anion 10000000 9000000

Performance Index

8000000 7000000 6000000

BETNH BTNH DMNH HETNH MNH ENH PNH C 8CHIN

B

Fig. 9. Performance index values at infinite dilution and at ambient temperature (T = 25 8C) for (A) pyrrolidinium and (B) phosphonium based IL with DDT.

that the [TBP], [C8CHIN], [BEPYR], [C4DMIM] and [C8MPY] have higher PI values for long carbon chain structure of EDCs. This result leads to the fact that the structural characteristics of ILs play a vital role in the extraction of both PCP and DDT. It can be seen that [TBP] cation is the most effective cation for the removal PCP and DDT. It provides an awesome indicator in the judicious choice of cations and anions.

5000000 4000000 3000000

3.9. Sigma profile

2000000 1000000

[OAc][B(CN)4][BF4][BMA][BOB][BTA][CF3SO3][CH3SO3][CH3SO4][C2H5SO4][C8H17SO4][Cl][DMPO4][HSO4][MAcA][MDEGSO4][N(CN)2][PF6][SCN][TOS][Sal][NO3][CF3COO][DEP][DEC][DBP][Br][DHPO4][ClO4]-

0

Anion Fig. 8. Performance index values at infinite dilution and at ambient temperature (T = 25 8C) for (A) phosphonium and (B) ammonium based IL with PCP.

In the COSMO system, sigma profile is the only descriptor which describes the local polarity of molecular surface and determines the interaction energies between the IL and extractant. For the choice of the ILs based on their selectivity values it is necessary to study the effect of hydrogen bonding on the extraction process. Among the imidazolium group ILs: [C8MIM][PF6] and [C4DMIM][PF6] were chosen for sigma profile calculations. Fig. 10 shows the sigma profiles of the ILs along with the sigma profile of DDT and PCP with water respectively. From Fig. 10(A) it

S.R. Pilli et al. / Journal of Industrial and Engineering Chemistry 18 (2012) 1983–1996

0.16

Water DDT [C4DMIM][PF6] [C8MPY][PF6]

0.14 0.12

A

p(σ)

0.10 0.08 0.06 0.04 0.02 0.00 -0.03

-0.02

-0.01

0.00

0.01

0.02

0.03

0.04

2

σ(e/Å ) 0.16

Water PCP [C8MIM][PF6] [C4DMIM[[PF6]

0.14 0.12

B

p(σ)

0.10 0.08 0.06 0.04 0.02 0.00 -0.03

-0.02

-0.01

0.00

0.01

0.02

0.03

0.04

2

σ(e/Å ) Fig. 10. Sigma profiles for (A) [C4DMIM][PF6], [C8MPY][PF6], DDT and Water, (B) [C8MIM][PF6], [C4DMIM][PF6], PCP and water.

was found that the sigma profile for [C4DMIM][PF6] and [C8MPY][PF6] were complimentary with the sigma profile of DDT. It should be noted that complementary sigma profile indicates higher solubility with each other. Thus DDT is highly soluble in both the ILs. The same phenomena is observed in Fig. 10(B) with profiles of [C8MIM][PF6] and [C4DMIM][PF6] with PCP. This again confirmed that the ILs are highly soluble in PCP. Both the compounds viz. effluent and IL have a higher fraction of donor as well as acceptor segments, thus the hydrogen bond is strong between compounds. CH–p interaction will mainly occur within segments, that are lying in between the region, i.e., shb (0.0082 e/A˚2) > s > + shb (0.0082 e/A˚2). These regions will be non-polar in nature and, thus, contribute mainly to CH–p interaction. In the sigma profiles of both DDT and ionic liquid, it is clear that both donor and acceptor segments are very small; thus, hydrogen bonding is less in comparison to PCP. The sigma profiles of PCP and DDT has a small fraction of hydrogen bond donor segments, which is a bit greater than hydrogen bond acceptor segments. 4. Conclusion Results of the present study showed that the extraction of endocrine disrupting organic effluents of environmental interest such as PCP and DDT has been screened theoretically. A screening of 35 cations and 29 anions, a total of 1015 possible ILs

1995

combinations was investigated via COSMO-RS based model to determine the best extractant for the removal of PCP and DDT from aqueous solution. Five different groups of cations namely imidazolium, pyridinium, pyrrolidinium, ammonium and phosphonium were investigated. All the selectivity calculations were done using infinite dilution activity coefficients as an indicator for the screening potential of ILs. In case of phosphonium based cations [TBP] gave high selectivity with anions [PF6], whereas [BOB] was the highest selective among all cations. Then the pyrrolidinium based ILs [DPPYR][C8H17SO4] and [BEPYR][ C8H17SO4] gave highest selectivity. In case of ammonium based ILs the highest selectivity values of PCP were observed as [C8CHIN][MAcA] (856) and [C8CHIN][CF3COO] (823) and other than cation [C8CHIN], selectivity order followed: [BTNH][Sal] (342), [BTNH][B(CN)4] (20.556). The selectivity order for the pyridinium based cations followed the order: [DPPYR] > [BEPYR] > [OMPYR] > [BMPYR] and for imidazolium based cations: [C8MIM] > [C4DMIM] > [C2DMIM] > [C6DMIM]. Similar trends were also seen for pyridinium cations. Highest selective pattern of different cations groups for the removal of PCP followed the order: phosphonium > pyrrolidinium > ammonium > Pyridinium > imidazolium. Anions such as [C8H17SO4], [MAcA], [Sal], [B(CN)4], [N(CN)2], [BTA], [CF3SO3], [MDEGSO4], [TOS] and [NO3] gave high selectivities because the sterical shielding effect around their charge centers was absent. Sigma profiles confirmed that the hydrogen bonding effect was weak in comparison to CH–p interaction of the effluent and IL. Benchmarking studies were carried for comparison of experimental data with predicted data for both PCP and DDT where the predictions agree well. Other 13 PCBs were also used for the validation of experimental data with theoretical data. Trends obtained in both the cases were similar to each other. The results of this study are encouraging and suggest that ILs can be a potential substitute for the treatment of organic rich water and wastewater. The accuracy of results obtained by the present method can be useful for further experimental investigations. Acknowledgments KM gratefully acknowledges that this work was financially supported by the research grant under the Fast Track Scheme from the Department of Science and Technology (DST), Government of India vide sanction No. SR/FTP/ETA-56/2010. References [1] L.D. Carlson, I. Basu, R.A. Hites, Environmental Science and Technology 38 (2004) 5290. [2] O. Wurl, R.P. John, P.O. Jeffrey, C. Durville, Environmental Science and Technology 40 (2006) 1454. [3] J. Fan, Y. Fan, Y. Pei, K. Wu, J. Wang, M. Fan, Separation and Purification Technology 61 (2008) 324. [4] H.S. Chang, K.H. Choo, B. Lee, S.J. Choi, Journal of Hazardous Materials 172 (2009) 1. [5] Access Science, McGraw–Hill Companies, New York, http://www.accessscience.com (accessed online 27.04.11). [6] S. Barlow, R.J. Kavlock, J.A. Moore, S.L. Schantz, D.M. Sheehan, D.L. Shuey, J.M. Lary, Teratology 60 (1999) 365. [7] D.L. Davis, H.L. Bradlow, M. Wolff, T. Woodruff, D.G. Hoel, H.A. Culver, Environmental Health Perspectives 101 (1993) 372. [8] N.R. Adams, Journal of Animal Science 73 (1995) 1509. [9] A.M. Comerton, R.C. Andrews, D.M. Bagley, P. Yang, Journal of Membrane Science 303 (2007) 267. [10] B. Ning, N. Graham, Y. Zhang, M. Nakonechny, E.G. Mohamed, Science and Engineering 29 (2007) 153. [11] S.R. Pilli, V.V. Goud, K. Mohanty, Environmental Engineering and Management Journal 9 (2010) 1715. [12] T.C. Zhang, S.C. Emary, Environmental Engineering Science 16 (1999) 417. [13] T.A. Ternes, M. Meisenheimer, D. McDowell, F. Sacher, H.J. Brauch, B.H. Gulde, G. Preuss, U. Wilme, N.Z. Seibert, Environmental Science and Technology 36 (2002) 3855. [14] J. Xu, L. Wu, W. Chen, A.C. Chang, Journal of Chromatography A 1202 (2008) 189.

1996

S.R. Pilli et al. / Journal of Industrial and Engineering Chemistry 18 (2012) 1983–1996

[15] M. Claraa, B. Strenna, O. Gans, E. Martinez, N. Kreuzinger, H. Kroiss, Water Research 39 (2005) 4797. [16] C. Susan, K. Wilson, C. Jones, Environmental Pollution 81 (1993) 229. [17] E.L. Ang, H. Zhao, P.J. Obbard, Enzyme and Microbial Technology 37 (2005) 487. [18] S. Sato, A. Matsumura, Journal of the Japan Petroleum Institute 46 (2003) 375. [19] A. Klamt, G. Schuu¨u¨rmann, Journal of the Chemical Society (1993) 799, II. [20] A. Klamt, Journal of Physical Chemistry 99 (1995) 2224. [21] F. Brett, Journal of Pesticide Reform 11 (1991) 1. [22] US Environmental Protection Agency, Pentachlorophenol Health Advisory, Office of Drinking Water, Washington, DC, 1987. [23] World Health Organization, DDT and its Derivatives in Drinking-water, Background document for development of WHO Guidelines for Drinking-water Quality, 2004. [24] G. Schaftenaar, J.H. Noordik, Journal of Computer-Aided Molecular Design 14 (2000) 123. [25] Frisch, et al., GAUSSIAN 03, Revision C.01, Gaussian, Inc, Pittsburgh, PA, 2003. [26] V.Z. Fock, Physik 61 (1930) 26. [27] A. Schaefer, C. Huber, R. Ahlrichs, Journal of Chemical Physics 100 (1994) 5829. [28] C. Sosa, J. Andzelm, B.C. Elkin, E. Wimmer, K.D. Dobbs, D.A. Dixon, Journal of Physical Chemistry A 96 (1992) 6630. [29] T. Banerjee, K.K. Verma, A. Khanna, AIChE Journal 54 (2008) 874.

[30] T. Banerjee, R.K. Sahoo, A. Khanna, Industrial and Engineering Chemistry Research 46 (2007) 1292. [31] M. Mukhopadhyay, B.D.M. Rao, Industrial and Engineering Chemistry Research XXIX (1987) T89. [32] S. Lin, S.I. Sandler, Industrial and Engineering Chemistry Research 41 (2002) 899. [33] R. Anantharaj, T. Banerjee, Fuel Processing Technology 92 (2011) 39. [34] R.M. Maduro, M. Aznar, Fluid Phase Equilibria 265 (2008) 129. [35] A.B. Pereiro, A. Rodriguez, Fluid Phase Equilibria 270 (2008) 23. [36] D. Mackay, Environmental Science and Technology 16 (1982) 274. [37] G.D. Veith, D.L. De, B.V. Foe, Bergstedt, Journal of the Fisheries Research Board of Canada 36 (1979) 1040. [38] W.Y. Shiu, K.C. Ma, D. Varhanickova, D. Mackay, Chemosphere 29 (1994) 1155. [39] P.G. Wightman, J.B. Fein, Applied Geochemistry 14 (1999) 319. [40] S.K. Park, A.R. Bielefeldt, Water Research 37 (2003) 3412. [41] W.H. Darryl, W.C. Des, Environmental Science and Technology 22 (1988) 382. [42] R.P. Gerber, P.S. Rafael de, Industrial and Engineering Chemistry Research 49 (2010) 7488. [43] R. Putnam, R. Taylor, A. Klamt, F. Eckert, M. Schiller, Industrial and Engineering Chemistry Research 42 (2003) 3635. [44] L.A. Blanchard, D. Hancu, E.J. Beckman, J.F. Brennecke, Nature 399 (1999) 28.