Indian Journal of Chemistry Vol. 42A, September 2003, pp. 2439-2446 Selective extraction of alkali metal ions from bittern using picrate anion and crown ethers as ligand: An experimental and theoretical study Pragati Agnihotri, Bishwajit Ganguly, E Suresh, Pari mal Paul* & Pushpito K Ghosh * Analytical Science Discipline, Central Salt and Marine Chemicals Research Institute, Bhavnagar 364 002, India Received 7 Janua ry 2003 Dibenzo-2 1-crown-7 (DB2IC7) and dibenzo-24-crown-8 (DB24C8) have been used to extract the metal io ns (Na+. K+. Mg2+ and Ca2+) present in bittern, schoenite and in an artificial solution containing equimolar amount of th e above mentioned metal ions. Ion chromatographic study of the organic extract shows that DB21 C7 extracts potassium with hi gh selectivity, whereas DB24C8 extracts comparable amount of potassium and sodium. The extraction eq uilibrium co nstants (Ke) for potassium and sodium with both the ligands have been evaluated spectrophotometrically by two-phase (wate rdichloromethane) extraction process. For DB2IC7, K. value for potassium is 4.5x104 M 2. which is 6.3x 102 time hi gher than that of sodium, whereas DB24C8 shows K. values 3.3xl04 and 2.4x104 M 2 for potassium and sod ium, respec ti ve ly. Molecular mechanics calculations have been performed to examine the selectivity of the cations. In the gas phase calculations, DB21 C7 and DB24C8 are found to be selective towards the sodi um ion . However, K+ is preferred by DB21 C7 in the aqueous solution, which is in qualitative agreement with our experimental results. Introduction The selective recognition and extraction of alkali metal ions is currently an area of great interest l -6 . Among various alkali metal ions, potassium ion recognition attracts particular attention because of its 7 wide range of applications , and also for elucidation of the potassium channel in the biological systems 8 . Bittern (mother liquor obtained after recovery of common salt) contains mainly Na+, K+, Mg2+, and small amount of Ca2+. The anions are mainly chloride and sulphate. It is one of the largest natural sources of K+, but not much attention has been given to the recovery of K+ from bittern through selective ex traction. We were interested to study the interaction of bittern with various ligands, which can selectively pull out K+ in presence of other ions in bittern. After the di scovery of crown ethers by Pedersen 9- 12 , these macrocycJic systems have been intensively l studied during the last three decades -5. MacrocycJic li gands are noted for their remarkable selectivity toward cations and anions making them excellent choices for the separation of these ions from mixtures l -3. 13 . It has generally been found that the selectivity depends on several factors , such as macrocycJjc cavity dimension , shape and topology, substituent effects, conformational flexibilitylrigidity and solvation effect. However, relative sizes of the cavity of the li gand and the ionic diameter of cations remain critically important factors that control the selectivity towards different ions 2,14. It has also been reported that flexible macrocycl es such as the large poly ethers, discriminate principall y among smaller cations (plateau selectivity)1 5.16. Among the cations present in bittern, the radius of K+ is larger than th ose of Na+, Mg2+ and Ca2+. Therefore, K+ ion should be the best candidate among th e ions for complexati on with the large macrocycJic li gands. We have studi ed this with dibenzo- crowns of 21-30-rin g size and th e obtained with dibenzo-30-crown -IO results (DB30ClO) have been submitted for publication l7 . We report herein the selectivity studi es for K+, a+. Mg2+ and Ca2+ cations in bittern , us ing dibenzo-21crown-7 (DB21 C7) and dibenzo-24-c rown-8 (DB24C8) as ligand. Studies were also conducted o n the aqueous by-product obtained upon decompos iti on of kainite-type (KCl.MgS0 4.3H 20 ) mixed sa lt into schoenite, and in an artificial mixture contain ing equimolar concentration of Na+, K+, Mg2+ and Ca ~+ for comparison. We also ' report the equilibrium constants for two-phase extraction of K+ and a+ using above-mentioned li gands and picrate ani on. Selectivity of the metal ions with DB21 C7 and DB24C8 has also been stud ied by mo lecular mechanics calculations using Monte Carlo search method. Materials and Methods The compounds dibenzo-21-crow n-7 , dibenzo-24crown-8 were purchased from Aldrich. Analyti ca l grade salts, KCI , NaCI , CaCI 2 , MgC I2 , KOH and 2440 INDIAN J CHEM , SEC A, SEPTEMBER 2003 NaOH were purchased from S.d.Fine Picric acid was procured from Sara Fine Analytical grade dichloromethane was from Ranbaxy and water was purified equipment Milli-Q, made by Millipore. Chemicals. Chemicals. purchased using the r rtracfion procedure Metal ions were extracted from aqueous solutions I y two-phase extraction procedure. The solutions f rom which extractions were made are: i) bittern, which contains Na+ (0.5 M) , K+ (0.7 M), Mg2+ (4.0 M) 2 and Ca + (0.1 M); ii) a by-product of schoenite manufacture, which contains Na+ (I.S M), K+ (2.0 M), Mg2+ (2.2 M) and trace amount of Ca 2+; and iii) an artificial solution containing Na+, K+, Mg2+ and Ca 2+, 0.1 M each . In a typical experiment, to a 5 mL solution was added picric acid 2 .3 mg (to make 2x lO·3 M solution) and 5 mL dichloromethane containing DB21C7IDB24CS (2x I0<1 M). The two-phase mixture was vigorously shaken on a vortex mixture for 10 min and the solution was transferred to a separating funnel and allowed to stand for 30 min. The clear dichloromethane layer was separated and passed through a bed of anhydrous MgS04 to remove trace amount of water. Solvent from the extract was removed and the solid product was isolated for selectivity studies. Selectivity defennination The solid product, obtained by the aboveme ntioned procedure, was taken in a crucible and heated in a furnace at 550°C for 4 h. The residue was di ssolved in deionised water (-5 mL) and filtered using 0 .2 11m filter paper. The concentrations of the cations in the filtrate were determined by ion chromatograph using Ion Pac CS 12 (2 mm) analytical column and 20 mM methylsulphonic acid as eluent with a flow rate of 0 .25 mUmin. Quantification was made using a standard solution containing a mixture of Na+, K+, Mg2+ and Ca2+ (-15 ppm each). A well resolved chromatogram was obtained, from which the amount of the cations in ppm/percentage were estimated using the software provided with the Dionex 500 ion chromatograph. solution of MOH (I X 10.3 M), (M = Na+ and K+) and 4 picric acid (I x I 0. M) was added 5 mL dichloromethane solution of DB21 C7/DB24C R (2xlO' s M). The two-phase reaction mixture was shaken vigorously and the organic phase was separated following the same procedure as described for extraction process. The conce ntrati o n of picrate in organic phase, which is equivalent to the metal ion extracted, was determined spectro photometricall y (Amax , 376 nm) . For a particular conce ntration of meta l ion, ten experiments were carried out with 0. 2 , OA . 0.6, O.S, 1.0, 1.2, 1.4, 1.S and 2 .0x I 0-4 M solutions of crown ether. Again, each set of ten ex periments was performed with I X 10-3 , IxIO-2 and Ix IO-' M concentrations of metal ion. Total th irty experiments were carried out for a particular me tal ion and a ligand, and the whole data has been used to calculate equilibrium constant for extraction (Ke) by non -lin ear regression method. Determination of partition coefficiellt Partition coeffi c ient of the crown ether between dichloromethane and water was determin ed by equilibration of the two phases, followed by the analysis of the aqueous layer by picrate extraction. In a typical experiment, 5 mL water was added to an equal volume of dichloromethane solution of DB21C7/DB24CS and the two phases were vigorously shaken (10 min), and aqueous layer was separated. To the aqueous layer was added solid KOH 1 (to make 0.1 M solution), picric acid (to make 5x I 0M solution) and an equal volume of dichloromethan e and the extraction was made following the same procedure as described above. The spectrophotometric analysis of the organic extract gave the concentration of complexed crown ethers. The concentration of the remaining c rown ether in dichloromethane phase was a lso checked spectrophotometrically, and it was close to the expected values. Partition coefficient was calculated using the equation Pe = [Cr]aq![Cr]org. This experiment was repeated with 5xlO-4, IxlO-3 , 4x lO-3 , SxIO-3 and IxlO-2 M dichloromethane solution of crown ethers. The average of the five values was considered for the calculation of equilibrium constant for extraction. Equilibrium constant for extraction Equilibrium constants for extraction were determined by measuring the concentration of picrate in the dichloromethane extract, obtained by two-phase extraction with various concentrations of crown and metal ion. In a typical experiment, to a 5 mL aqueous Stability constant in the aqueous phase The stability constants for complex formation in the aqueous phase were determined by the following experiments. To the 5 mL aqueous solutions of KOH (l X 10-3, 5xlO-3 , IxlO- 2, 5xlO-2, and Ix IO-' M) was AGNIHOTRI et al.: SELECTIVE EXTRACTION OF ALKALI METAL IONS FROM BITTERN added required amount of 0821 C7/0824C8 to make I x w - ~ M concentration and the reaction mixtures were stirred for 24 h at room temperature. The solutions were then filtered through a 0 .2 J.UTI pore size filter paper and the absorbance of the filtrate was measured spectrophotometrically at 274 nm C"-rnax)Owing to the complex formation the concentration of the ligand in solution increased with increase in the concentration of the metal ions. The solubility of the ligand under saturation condition was determined in the following way. The solid ligand (5.0 mg) was suspended in 50 mL of water and stirred for 16 h. After every 2 h interval, about 5 mL aliquot was taken, filtered through 0.2 11m pore size filter paper, and the concentration of the ligand in the filtrate was determined spectrophotometrically. This was continued until the solution became saturated, when there was no significant change in the absorbance of the 274 nm band. The average of the three limiting values was considered as the concentration of the ligand in saturated solution. The UV-vis spectra were recorded on a model 8452A Hewlett-Packard diode array spectrophotometer. Cation concentrations were measured on a Oionex 500 ion chromatograph using Ion Pac CS12 (2 mm) analytical column. Results and Discussion Selectivity Selectivity of the metal ions (Na+, K+, Mg 2+ and Ca2+) was determined by two-phase extraction method usi ng crown ethers and picrate anion, followed by analysis of the concentration of cations in the extract by ion chromatography, as described in the ex perimental section. The chromatogram obtained by analysis of the extract from bittern using 0821C7 is shown in Fig. 1. The concentrations of metal ions in the original solutions (before extraction) and the selectivity ratio (K+/M"+; M"+ = Na+, Mg 2+, ci+) in the organic extract are shown in Table 1. The data suggest that 0821 C7 has significantly higher selectivity towards potassium compared to that of D824C8 . The artificial mixture, where all of the metal ions have equal concentration, and also schoenite by-product show K+/Na+ close to 6 for 0821 C7; whereas the corresponding ratios with D824C8 are 1.2 and 2.0, respectively. However, in bittern 0821C7 shows remarkably high selectivity towards potassi um. High Mg 2+ concentration and/or 2441 ECD 1 cationPaul170602 200 CATION PAUL 170602 #95 .iJS 100 E 1 _J I I 0. 1 4.0 6 .0 E .i! 2 :c Hc E ~ u go h ·:• : 2 ;.:~ A ~ . ' ' I 8 .0 10.0 "' 12.0 mu"'i I 14.0 16 0 -, ~- 18 0 .- . 20 0 Fig. l- Ion chromatogram showing the di stribution of metal io n' extracted from bittern usi ng DB21C7 and picrate anio n. other anions present in bittern might have so me effect on selectivity. Mg 2+ and Ca2+, in general, show hi gher K+/M 2+ values compared to Na+, which is mainl y because of their smaller size. High selectivity for potassium by 0821 C7 is related to the interaction energy of the hydrated cation complex and also low energy confo rmation of the li gand , which has been discussed in the theoretical studies section . Extraction equilibrium Two-phase extraction using picrate ani on has been used for quantitative determination of relati ve complexing ability of the li gands6 ' 18' 19 • The picrate anion has an absorption band at 378 nm in dichloromethane, whereas crown ethers do not absorb in this region. Therefore, the concentration of picrate, which is equivalent to the concentration of the univalent metal ion, can be determined spectrophotometrically. The extraction experiments were carried out in water-dichloro meth ane mi xture with various concentrations of li gands (2x I o-5 to 2xl0-4M) and metal ions (l x l0-3, l x i0-2 and lx i0-1 M), following similar procedure as described for selectivity studies. The UY -vi s spectral change of the organic phase for both the cations ( K+ and Na+) with increasing the concentrations of 0821 C7 is shown in Fig. 2. We also carried out blank experiments, where extraction was carried out under similar experime ntal conditions but without crown ethers, and no spectrophotometrically detectable amount of pi crate is observed in the organic phase. The equilibrium constant was calculated following the method publi shed by Frensdorff 18 . The two-phase equilibrium containing metal ion (M+), picrate ani on (A'), and crown ethers (CE) can be defined as : INDIAN J CHEM , SEC A, SEPTEMBER 2003 2442 Table !-Concentratio ns of metal ions in the original solutions and potassium selectivity rati os (K+/M"+) in the ex tract Soluti on Before extraction (M) [Na•] [K+] [Mg 2•] [Ca 1• ] Bittern 0.5 Schoenite 1.8 0.7 4.0 K•/N a• 0.1 2.2 2.0 In the extract DB21C7 K•/Mg2• K•tca 2• K•/N a• Mi xture" 0. 1 0.1 0. 1 0. 1 '' rrace amo unt of Ca 2 • is present tAn artifi cial solutio n containing equimolar amount of cations 15.7 18.3 6.2 22. 3 5.9 11.8 3 1.4 8. 5 K =[I+ P,. + P,.KJ(M 0 2.4 . ' 2f[M 0 - DB24CX K • /M g ~• 1.4 2.2 2.0 1.8 1.2 4.0 K • t ca~· 7.9 3.2 1 A)][2A+K" - (K,7- 4K,1 A) n ] A][A 0 - A][CE 0 - A] - . .. (5) 1.9 O.t t Q) u c 1.4 ro 220 320 42t 520 120 .0 '0 VI .0 <( 0 .9 0.4 -0 .1 320 220 420 520 620 Wavelength (nm) Fig. 2-UV -v is spectral changes in the organic phase with increasing the concentrati on of DB2 1C7 (2x l0·5 to 2x!04 ) using 0.1 M K+ and I x I 04 M picrate. Inset shows spectral changes under similar conditio ns with 0.1 M Na•. where Kd is the di ssociation constant of MC EA in the organic phase, Pe is the partition coeffi cient of the ligands in water-dichloromethane, /(, is the stab ility constant of MCE+ in aqueous phase, f is the single ion activity coefficient of the aqueous cation , A is th e concentration of the picrate anion in organic phase: Mo. A0 and CEo are the initial concentrations of th e metal ion, picrate and crown, respectively . The extraction equilibrium constant Ke can be determ ined from Eqn (5) by non-linear regression fitting provided Pe and K_, are known . The partition coefficients <Pe) for both the li gand s were determined following the method described in the experimental section . The stability constant for complex form ation in the aqueous phase (K,) was also determined spectrophotometrically followin g a recently published procedure20 '2 1, which is suitable for nearly insoluble (in water) ligands. The stabi lity constant can be determined using Eqn (6) . . . . (1) .. . (6) where MCEAarg designates ion pair. To evaluate Ke, the distribution of all the species in both phases has to be considered, which can be accounted by the following equilibriums. Kd MCEAors~ CE org~ MCE:,g + A ~ , g ... (2) Pe . . . (3) CEaq M;q + CEaq ~ Ks MCE;q . .. (4) From Eqns (1) to (4), the final equation for Ke can be expressed as 18 ' 19 : where A is the absorbance due to ligand and complex formed, Aa is the absorbance due to ligand under saturated condition, E 1 are E 2 are molar absorptivities of the ligand and its metal-complex, respectivel y: [CE]sat is the concentration of the ligand, and C,. 11 is the initial concentration of the metal ion. A and A.. can be determined following the methods described in the experimental section. Plotting N A0 -l as a function of Csat 1; and applying the conditions: i) the value of t 2 and t 1 are very close (e2/t 1==1), and ii) solubility of the ligand in water is very low ( b[CrL 11 << 1), the slope of the straight line gives K} 0 ·21 . AGNIHOTRI et a/.: SELECTIVE EXTR ACTION OF ALK ALI M ET AL IO NS FROM BITIERN 2 4 4~ Table 2- Partiti on coeffi cients and equilibrium constants in volving ex trac ti on process Cati on Pe x I 0 3 K,, M-1 Kd x I 0 5 , M K+ 6.27 15.9 84.0 6.27 2.3 0. 1 0. 11 8.9 1.2 3.3 ± 0.4 X 1 0~ DB24C8 Na• 0. 11 8. 1 1.0 " Because of the small changes in the UV -vis spec trum the standard deviation is too hi gh 2.4 ± 0. 2 X 104 Li gand D82 1C7 DB2 1C7 DB24C8 K+ Putting all the experime ntal data in Eqn (5), re maining unknown s in thi s equation were Ke and K". The value of Ke was e valuated by non-linear regress ion analysis using all the 30 data set with vari ed K" to minimi ze the standard deviation among the calcul ated values. The Ke values along with Kd, K., and Pe are show n in Table 2. A pl ot for the fraction of co mpl exed crown extracted as a function of li gand concentration fo r OB 2 1C7 with K+ is shown in Fi g. 3. The ex tracti on equilibrium constant describes qu antitatively the effici e ncy of extraction from aqueous to organi c phase. It reflects not only the strength of the co mpl ex formed but al so ability of making effect ive ion pair between complex-cati on and anioni c species, and also solubility of the vari ous spec ies formed. The picrate anio n makes strong n-n stacking interacti ons wi th the benzene rings of the dibe nzo-crown ethers, in additi on the oxygen ato ms of the pi crate ani on also interact strongly with the crow n mo iety 22 . The co mbinati ons of these two effects make strong io n pair, whi ch facilitates the ex tracti on process 22 . In fac t, we also observed strong n-n stackin g and 0 . . . H interacti ons between pi c rate and crow n-cati on mo iety in the crystal structure of [K-DB 30C lO] [pi crate] 17 . T he data in T abl e 2 shows th at the Ke valu e fo r OB 2 1C7 with K+ is re markabl y 2 high co mpared to th at w ith Na• (6.2x l0 time). Howeve r, for D8 24C8 the Ke values fo r K+ and Na+ are co mparabl e, indi cating th at 082 1C7 is a much better ex tractant fo r selective ex tracti o n of potass ium . Th eoretical studies A number of theoret ical studi es including mo lec ul ar mecha ni cs, molecul ar dyna mi cs, Mo nte Carl o simul ati ons and qu antum che mi cal calcul ati o ns have been perfor med on macrocycl ic receptor molecul es such as crow n e the rs, c ryptands and he mispherand s23 -25 . In gene ral, the calcul ated res ul ts have quali tative ly reprod uced the ex peri mental tre nds. In macrocyc li c che mistry, molec ul ar mech a ni cs has 2 K,, M- 4.5 ± 0.5x I0~ 72.0 ± 16.0" 0.5 O.l M • 0 0.4 ~ "0 <ll ura x'-<ll c: • 0.3 0 .2 O.Ol M 0 ura '- u_ 0 .1 :-~ 0 0 ~ M 0.5 1.5 2 2.5 CroWl :picrate (CEo/Ao) Fig. 3- Piot fo r the frac ti on of co mple xed crow n ex trac ted as a fun cti on of liga nd concentrati o n (2x I o-5 to 2x I o - ~) with va ri ous metal ion co ncentratios. gained importance as calc ulati ons are less expe ns ive co mpared to qu antum che mi cal meth ods and th e 21 results obtained from MM methods are also reliabl c . Th e co mpl exati on study of mixed alkali and alkalin e eart h metal 1.o ns e.g., Na +, K+, Mg 2+ and a-'+ prese nt in bittern , however, with 08 2 1C7 a nd D8 24C8 has not been performed either experimentally or We have e mpl oyed mo lecular theoreti call y. mechani cs merck fo rce fi eld (M MFF94) usin g Mont e Carl o confo rm ati on search method to e xami ne the conform ati ons of 08 2 1C7 and O B24C8 and its co mpl exati on w ith the alkali metal ions prese nt in bittern 26 . The con fo rmati o ns obtained fro m th e searc h fo r 08 2 1C7 and D8 24C8 has bee n used for furt he r co mpl exati o n study with th e alka li metal io ns. To in vesti gate the ro le of aqueous so lve nt , we exa mine th e io ni c excha nge reac ti on. c ~ K+(OB 2 1C7/D8 24C8) + M +(H 20 ), K+(l-h O ), + M\ D82 1C7/D8 24C8 ) . .. (7) INDIAN J CHEM , SEC A, SEPTEMBER 2003 2444 for which the metal cation (M+ = Na) coord in ated by n = 6 molecules of hydration, displaces a K+ cation from 0821 C7 or OB24C8. The exothermicity of this reaction can be considered a measure of selectivity for the cations. Wipff et a!. previous ly reported the se lectivity of alkali meta l ions for Eqn (I) and n = 6 · ustn g molecular mec h antes approac h 23ab· . onetheless, high level quantum chemical calculations have also been performed with a fewer Ta ble ] - interaction energie s of cati o n-hexahydrate comp lexes molecules of hydration for predicting the se lecti vity of metal ions a nd the results could qualitative ly reproduce the experimental trend 2:1i.o.q.' . The calculated interaction energies of cation hexahydrate co mplexes at molecular mechanics level are shown in Tabl e l . Sodium and potassium ions show a very good agreement between experiment and theory. A total of 2,000 to 2,500 trial conformation s were generated for 082 I C7 and OB24C8 of which I 00 were retained in each case based on structural dissimilarity and energetic considerations. So me of the low energy conformations identified for 0821 C7 and OB24C8 are shown in Figs 4a and Sa. Th e ethyle ne -oxy- units at each side of th e ben ze ne rin g can rotate and generate the major structural changes fou nd in these set of conformers . The frequentl y sampled conformations found in both th e cases are Co mpl ex Na+. . ... ( OH ~ ) <> 100.2 96.4 8 1.3 79. 7 K+. . ... . (OH 2) 6 ·' Total energy of co mpl ex= total interaction e nerg y (kcal/mo l) " Ex perim ental hexahydrati o n enthalpy (re f. 27) (a) M:=O.O (I) t.E=1.8 (Ill ) t.E=0.6 (II) M:=2.8 (IV) (b) ~-¥f;t! t.E=O.O (I) t.E=O. I (II) ~ t.E=l.O (Ill) ~ M:= l.S (IV) (c) M:=O.O (I) M:=0.2 (ll) t.E=O.S (Ill) M:=0.7 (IV) M:=2.0 (V) Fig . 4-Low energy confo rmati o ns of (a) un comp lexed DB21 C7 , (b) Na +-DB2 1C7 co mpl ex, and (c) K+-DB2 IC7 co mplex (re lati ve energ ies in kcal/mol ) AGNIHOTRI el c~l.: SELECTIVE EXTRACTION OF ALKALI METAL IONS FROM BITTERN 2445 (a) t..E=O.O t..E=0 .8 (I) (II) L'lE=L8 (IV) t..E=U (III) (b) L'lE=O.O . (I) L'lE= LO (II ) L'lE= l.7 (Ill) (c) L'lE=O.O (I) t..E=O . 1 (II) L'.E=LO (III) Fig. 5- Low energy conformatio ns of (a) uncomplexed DB24C28, (b) Na+-DB24C8 co mplex. and (c) K+-DB24C8 complex (relati ve energies in kcal/mol ) unfolded forms of ligands DB21C7 and DB24C8 (Figs 4a and 5a). The lowest e nergy conformers Figs [4a(i)] and [5a(i)] were taken for the complexation study with alkali metal ions Na+ and K+. The metal ions were located near the center of the conformation for further calcul ati ons. In each case, around 2000 to 2500 co nformers have been sampled and I 00 conformers were reta ined . The low energy samp led conformations for the co mp lexation of 0821 C7 and D824C8 with Na+ and K+ are shown in Figs 4b,c and 5b,c, respectively. The calculated results suggest that the ligand 0821 C7 wraps around the sodi um ion while complexing (Fig. 4b), whereas the potassi um ion prefers to sit at the periphery of ligand (Fig. 4c). The ligand 0821 C7 may not be ab le to wrap around the K+ due to its relatively large diameter (2.66 A) than that of Na+ ( 1.90 A) . However, the relatively larger ligand D824C8 wraps aro und both the Na+ and K+ completely (Figs Sb and c). ln the gas phase, both the li gands 08 2 1C7 and DB24C8 prefer to bind selectively with th e sodium ion. The calculated binding enthalpy suggest that the sodium ion binds 17 .5 kcallmol stro nger than th e corres pondin g potassium ion. The gas phase binding en th alpy, calculated for the complexation of sodium ion with D8 24C8 is even stron ger (21.3 kcal/mo l) than that of potassium ion. However, by incorporatin g the waters of hydration (n = 6) as suggested in Eq . 7 the se lectivity changes for the li gand DB21 C7. Potassium ion compl exed with 0821 C7 has been found to be energeti ca lly more stable compared to sod ium ion in the aqueous phase by 1.9 kcal / mo l in qualitative agreement with our experimental results. In the case of D824C8 , sodium ion has bee n found more stable than that of potassium ion by 1.8 kcallmo l in the aqueous phase calcu lation. Our experime ntal results also suggest that the ligand D824C8 has comparable affin ity for both sodium and 'potassium ions. It is worth -men ti oning that in the mixture o f 2446 INDIAN J CHEM, SEC A. SEPTEMBER 2003 vanous ion s like bittern, competitive kinetics must also play a ro le towards the selectivity of metal ion s. 13 14 Conclusion We have examined the cation selectivity using 0821 C7 and D824C8 in bittern, schoen ite byproduct liquor, and a n artificial mixture containing equi molar amount of cations present in bittern. The se lectivity study by ion chromatography show s that 08 2 1C7 extracts potassium with very high se lectivity in presence of other cations in bittern. The two-phase extraction equilibrium constant for potassium with 0821 C7 is remarkably high compared to that of sod ium. However, DB24C8 did not show much affinity for potassium over sodium, as evident from selectiv ity study and extraction equi librium constants. The calculated results suggest that the ligands 0821C7 and OB24C8 are selective towards sodium io n in the gas phase. However, the aqueous phase calc ulation s show the preference for the potassium ion with DB21C7 in qualitative agreement with our experimen tal observa tion s. OB24C8 in the aqueous media, however, shows slight preference for sodium over potassium . Experimentally also we observed that this li gand has almos t equal affinity for potassium and sodium ions. Therefore, the specificity appears to be a de licate balance between the energies of crown-cation interactions and cati o n so lvation . Ackno~ 15 16 17 18 19 20 21 22 23 · ~edgmnt Fi nancia l su pport from the Department of Science and Technology (OST), Government of India, is grateful! y acknow Iedged. References !zan R M, Brad,:1:1w J S. Nielsen S A, Lamb J D & Chri stensen J J. Chem He\', 85 ( 1985) 27 1. 2 Baj:.1j A V & Paonia N S, Coord Chem Rev, 87 ( 1988) 55. 3 !zan R M, Pawlak K & Bradshaw J S, Chem Rev, 91 (1991 ) 172 i. 4 5 Zhang X X, !zan R M, Bradshaw J S & Krakowiak K E, Coord Chem Rev, 174 (1998) 179. Buhlmann P, Pretsch E & Bakker E. Chem Rev, 98 (1998) 24 69 . 9 I0 II 25 26 Pedersen C J. JAm chem Soc, 89 ( 1967) 7017. Pedersen C J, JAm chem Soc, 92 ( 1970) 391. Pedersen C J & Frensdorf H K, Angew Chem lnt Edit. II ( 1972) 16. 12 Pedersen C J. Angew Chem lnt Ed Engl. 27 ( 1988) I 021. Leh J-M , Science, 227 (1985) 849. Izatt R M & Christensen J J. Synthesis o( 1/WtTon/es. '!h i' designin g of selective complexing agen1.1 . (Wile) lnterscience, New York), ( 1987). Vogtle F & Weber E. in Crown eth ers and wwlugs. edited b.1 S Patai & Z Rappoport, (Wiley, New York). ( 1989 ) 207. Gootenhuis P D J. Wal P D V & Reinhoudt D N. Tetrahedron, 43 ( 1987) 397 . Ag nihotri P. Suresh E. Ganguly B. Paul P & Ghosh I' K. submined for publication. Frensdorff H K. JAm chem Soc. 93 ( 197 1) 4684. He G-X , Abe A, Ikeda T. Wada F, Kikukawa K & Mat,utb T, Bull chem Soc Japan , 59 ( 1986) 674. Buschmann H-J , Cleve E. Denter U & Scho llmeyer E. ./ ph1·.1 urg Ch em , I 0 ( 1997) 781. Buschmann. H-J. Weng G. Cle ve E & Schollmeyer E. ; Ida Chim Slov, 47 (2000) 55. :~ Jr R E. Talanova G G. Elkarim N S A, Ta lanov V S. Ham Hwang H-S, Bartsch R A & Rogers R D. ./ Am che111 Soc ·. 12 I ( 1999) I l 28 I. (a) Wipff G, Weiner P & Kollman P, JAm clle/11 Soc. 10-+ (1982) 3249: (b) Grootenhius P D J & Kollman P. J ,\m chem Soc. Ill (1989) 2 152; (c) Bovill M J. Chadwick D J. Sutherland I 0 & Watkin D, J chem Soc Perkin Tm11.1 2. ( 1980) 1529; (d) Dobler M, Chimia . 38 ( 1984) 415: (e) Drc '' M G B & Nicholson D G. J chem Soc Dalton Tm11.1. ( J9 X6) 1543; (f) Grootenhius P D J. Eerden J van, Sudh olter E J R. Reinhoudt D N, Roos A, Harke ma S & Fe il D. JAm cht'lll Soc , 109 ( 1987) 4792 ; (g) Damewood J R Jr. Urban J J. Wil li amson T C & Rheingold A L. J org Che111. 53 ( J9R7l 167 ; (h) Weiner P K, Profeta Jr S, WipiT Ci . Havel T. Kunt z I D, Langride R & Kollman P A. Tetmhetlron. Jl) ( 19X.'\ 1 111 3; (i) Feller D. Tho mpso n M A & Ke ndall R A. ./ Phn Chem A, 101 (1997) 7292; (j) Anderson W P. Bellm Jr J> &. Glennon T M, J phys Chem. I0 I ( 1997) 1920; (k) Dang L X & Kollman P A . ./ phys Chem, 99 ( 1995 ) 55: (I) Eerden J van, Harkema S & FeilD. J phys Chem . 92 ( 1988) 5076: ( 111 1 Ranghino G, Romano S. Lehn J-M & Wipff G. ./ Am che111 Soc, I 07 ( 1985) 7873: (n) Ciun steren W F va n & Bercnd ~c n H J C. in Molecular dynamics anti protein struc/un•. cdi tL'd by J Hermans, Ed: Polycrystal Book Sen·icc·: We,tcrn Sprin gs, {1985) 5: {o) Feller D. J phys Chem A. 101 ( 1997) 2723; (p) Anderso W P, Behm Jr P. Glennon T M & Zcrncr M, J phys Chem A. 101 (1997) 1920: (q) Glendenin g E IJ. Fel ler D & Thompso M, JAm chem Soc. 116 (1994) 1065 7: (r) Kollman P A, Grootenhuis P D J & Lopez M A. Pun· appl Chem. 6 1 ( 1989) 593 : (s) Glendening ED & Fel ler D. .I Am chem Soc. 118 (1996) 6052: (t) Straatsma T P & McCammon J A, J chem Phys, 91 ( 1989) 3o3 1. (a) Geue R, Jacobso n S H & Pizer R. J Am ch e111 Soc. IOX (1986) 1150, (b) Wipff G & Kollman P A. Nm •u J Chi111. 9 ( 1985) 457. 1593. Kim J S, Shon 0 J, Ko J W, Cho M H, Yu I Y & Vicens J, J Org Chem . 65 (2 000) 2392. 7 Krau ss A A. Ka/i und Steinsalz, 12 (1999) 12R. R Doyle D A, Cabral J M. Pfuetzner R A, Kuo A, Gulbis J M. Cohen S L. Chait B T & MacKinnon R, Science. 280 ( 1998) 6 ~ 27 (a) Grootenhuis PD. Eerden J va n. Dijk str P J. Hakcma S & Reinhoudt D N, ./ Am che111 Soc, 109 ( 1987 ) XO-+-+: (h) Kol lman P A, Wipff G & Singh U C. JAm che111 Soc. 107 ( 1985) 2212: (c) Venanzi C A & Bunce J D. lnt J quant Clwn, Quant Boil Symp, 12 ( 1986) 69. (a) Halgren T A, J Comput Chem. 17 ( 1996) 490: (b) TIT AN. Wave.fi.mction. In c, 1840 I Von Karman Avenue. Suite 370, Irvine CA 926 i 2 USA . Schrndinger. In c. I 500 SW First Avenue, Suite 1180, Portland. OR 9720 I USA Dzidic I & Kebarle P, J phys Che111 . 74 ( 1970) 1466. Sec also, Kebarle P, Am111 Rev Phys Chem. 28 ( 1977) -+45 .