ANALYST, APRIL 1989, VOL. 113 45 1 Enhanced Extraction of Lanthanides with Crown Ether Carboxylic Acids of Increasing Li po p h i licity Jian Tang and Chien M. Wai* Department of Chemistry, University of Idaho, Moscow, ID 83843, USA sym-Dibenzo-I 6-crown-5-oxyacetic acid and its modified analogues can extract lanthanides efficiently in solutions with high ionic strength and complex matrices. Increasing the length of the side-arm alkyl group increases the lipophilicity of the macrocyclic polyether and enhances the distribution ratio of the lanthanide complexes in the organic phase. This type of crown ether carboxylic acid is a potential complexing agent for seiective extraction and concentration of lanthanides in natural water systems. Keywords: Lanthanide; extraction; crown ether; lipophilicity Macrocyclic polyethers, as a new generation of specific complexing agents, have found many attractive applications in separation chemistry and in analytical chemistry.Selective complexation of cations with crown .ethers can be achieved based on the cation radius - cavity size compatibility concept. For the extraction of cations with neutral crown ethers, the nature of the counter anions involved is an important factor in determining the efficiency of the extraction. The role of the counter anions can be filled by introducing anionic functional groups to the structures of the crown ethers. -4 number of functionalised crown ethers have been synthesised and have been shown to extract alkali metal ions and alkaline earth- metal ions in aqueous solutions independent of counter anions.1.2 The combination of ion-binding cavities possessing fixed dimensions with ionisable functional groups may create novel bifunctional complexing agents with extraction effici- encies and selectivities surpassing the closely related non- ionisable crown ethers. Recent reports have indicated that some crown ethers with pendant carboxylate functional groups are surprisingly efficient and selective for extracting trivalent lanthanide ions.-'.-.' Because of their high selectivity and anion independence, these crown ether carboxylic acids may have potential analytical applications for lanthanide extraction from complex aqueous systems. Neutron activation analysis .(NAA) is one of the most sensitive methods for analysing the lanthanides.5 However, direct application of NAA to lanthanide determination in natural waters is often not possible because of the low lanthanide concentrations and the matrix interferences, espe- cially in complex systems such as sea water.Therefore. separation and concentration procedures are generally required for NAA determination of lanthanides in natural systems. Besides 'being selective for lanthanide extraction, crown ether carboxylic acids have another attractive aspect for NAA, i.e., they are inert to neutron irradiation and hence would not create spectral interferences. Recently we reported the complexation of a crown ether carboxylic acid. sym- dibenzo-16-crown-5-oxyacetic acid, with trivalent lanthanide ions in a de-ionised water system.4 This macrocyclic poly- ether, with an attached carboxylate group, is effective for lanthanide complexation but has a drawback due to its weak lipophilicity which causes slow phase separation and transport of the complex to the organic phase.The lipophilicity can be enhanced by adding alkyl groups to the attached side-arm. In this work we have studied the extraction efficiencies of several crown ether carboxylic acids with modified sidearm struc- tures in order to elucidate the factors controlling the efficiency of extraction. The effects of alkyl substitution and lipophilicity of this type of crown ether carboxylic acid on lanthanide * To whom correspondence should be addressed. extraction and their potential applications to lanthanide determination in solutions with high ionic strength and complex matrices are described in this paper.Experimental Reagents sym-Dibenzo-16-crown-5-oxyacetic acid (I), 2-(sym-dibenzo- 16-crown-5-oxy)hexanoic acid (11) and 2-(dibenzo-l6-cruwn- 5-oxy)stearic acid (111) (Fig. 1) were synthesised according to the procedures given in the literature.6 The basic structure of the macrocycle was synthesised by the reaction of catechol with bis(2-chlorethyl) ether to form bis[2-(o-hydroxy- phenoxy)ethyl] ether followed by its reaction with epichloro- hydrin to form syrn-5-hydroxydibenzo-16-crown-5. The pen- dant carboxylic acid group was introduced to the niacrocycle by its reaction with bromoacetic acid, 2-bromohexanoic acid and 2-bromostearic acid to form I, 11 and 111, respectively. Three lanthanides, La-i+, Eu'f and Lu-T+, were chosen to represent the lanthanide series.The lanthanides, in nitrate or in oxide forms, were obtained from Alfa Products. Other chemicals used were all Baker Analyzed Reagents. De- ionised water was prepared by passing distilled water through R A 0 0 0 0 Fig. 1. Structures of sym-dibenzo-16-crown-5-oxyacetic acid ( I ) , 2-(sym-dibenzo-16-crown-5-oxy)hexanoic acid (11) and 2-(sjr)z-di- benzo- 16-crown-5-oxy)stcaric acid (111)452 ANALYST, APRIL 1989. VOL. 114 an ion-exchange column (Barnstead Ultrapure Water Purifi- cation Cartridge) and a 0.2-pm filter assembly (Pall, Ultipor DFA). All containers were acid-washed, rinsed with de- ionised water and dried in a class 100 clean hood. A synthetic sea water which was prepared according to the literature7 was used to evaluate the conditions for lanthanide extraction.Extraction Procedure The extraction solutions were prepared by dissolving weighed amounts of the crown ether carboxylic acids in chloroform in beakers with magnetic stirring. After dissolution, the organic phase was shaken with an HCI solution at pH 2 to remove potential metal impurities in the system. After purification, the organic phase was kept in contact with an LiOH solution to maintain a pH level of about 7. In some experiments, radioisotopes were used as tracers to test the extraction efficiency. In other experiments, p.p.b. to sub-p.p.b. levels of La3+, E u ~ + and Lu3+ were added to water samples to study the recovery by NAA. The water samples were adjusted to a desirable pH with LiOH and acetic acid.In general. to each water sample (100 ml in a glass-stoppered flask), 10 ml of the extraction solution were added and the mixture was shaken vigorously on a mechanical wrist-action shaker (Burrell Model 75) for a fixed time (usually 2 min) at room temperature. After shaking, the mixture was allowed to stand for a few minutes for the phase separation to be completed. For the tracer experiments, 5 ml each of the organic and aqueous phases were pipetted out of the flask and placed in 10-ml glass vials with fast-turn caps for y counting. For NAA experiments, 5-8 ml of the organic phase were removed from the flask and placed in contact with 1.5 ml of HN03 solution at pH 2 in another flask. The mixture was shaken again for 2 min to back-extract the lanthanides into the acid solution.After phase separation, 0.5 ml of the acid solution were placed in a % dram (ca. 1.48 ml) polyethylene vial which was later heat sealed for neutron irradiation. Neutron Irradiation and Activation Analysis Sample irradiations were performed in a 1-MW TRIGA nuclear reactor with a steady neutron flux of 6 X 10’2 n cm-2 s-1. Samples were generally irradiated for 2 h followed by 1 d of cooling before counting. The half-lives of the isotopes produced and the y radiations used for their identification are given as follows: I-loLa 400.2 h, 487 keV; 152Eu”’ 9.3 h, 122 keV; and 1’7Lu 6.7 d, 208 keV. A large volume Ortec Ge(Li) detector with a resolution of about 2.3 keV at the 1332 keV y line from W o was used for y counting.Signals from the detector were connected to an EG&G Ortec ADCAM (Model 950A) multi-channel analyser with software and an IBM-PC for data processing. The details of sample irradia- tion, y counting and the general procedure of NAA are given elsewhere. 8 Results and Discussion Extraction of Lanthanides From Sea Water With sym-Dibenzo- 16-crown-5-oxyacetic Acid The efficiencies of extracting the lanthanides from synthetic sea water with I as a function of pH at an organic (chloroform) to aqueous volume ratio of 1 : 1 are shown in Fig. 2. For the extraction of 2 X 10-7 M Lu3+ from the synthetic sea water with 3 x 10-3 M I in an equal volume of the organic phase, the extent of extraction was about 95-96% at pH ca. 6. However, PH Fig. 2. water by the crown ether carhoxylic acids 1-111 pH dcpendence of the extraction of L u 3 from synthetic sea if the organic to aqueous phase ratio was decreased to 1 : 10, the extraction was found to be much lower, at about 65%.The distribution ratio ( D ) is defined by the following equation: Based on the percentages of extraction measured at different organic to aqueous phase ratios, the value of D for Lu3+ was calculated to be about 20. Similarly, the D values for E d + and La’+ were calculated to be 10 and 5 , respectively. Increasing the concentration of the chelating agent should enhance the percentage of extraction of the lanthanides. However, the weak lipophilic property of I, the slow phase separation and the possible micelle formation become an experimental problem when the concentration of the chelating agent becomes high.Another approach,is to use a multiple- extraction procedure by repeatedly extracting a water sample with the organic phase containing 3 X 10-3 M of the crown ether carboxylic acid. For the extraction of 2 x lo-’ M Lu-7+ from 100 ml of the synthetic sea water with triple extraction of 10 ml each of the extractant solution, a total extraction efficiency of about 92% for Lu3+ was obtained. The total recoveries for E u ~ + and La’+ after the triple extractions under the same experimental conditions were about 88 and 60%. respectively. The solubility of I in water at room temperature has been estimated to be about 5.8 x 10-4 M.’ It has also been shown that in a chloroform - water two-phase system, the concentra- tion of I in the organic phase declines rapidly as the p€I increases from 5 to 8 and then reaches a plateau for the region pH 8-12.In this latter region, only 20-30% of the initial concentration of I remains in the organic phase.’ When a butyl group is attached to the side-arm of I , the solubility of the resulting crown ether carboxylic acid I1 in water has been reported to decrease by about an order of magnitude relative to that of I.” Extraction with 2-(sym-Dibenzo-16-crown-5-oxy)hexanoic Acid Two important observations were made when I1 was applied to extract the lanthanides from the sea water: (1) the phase separation in this instance was much faster and (2) the distribution ratios for the lanthanides were much greater compared with I. Replacing H in the side-arm of I with C4H9 significantly improved the lipophilicity of the chelating agent and greatly facilitated the phase separation process.Accord- ing to Strzelbicki and Bartsch,v in a chloroform - water system, the concentration of I1 in the organic phase remains virtually unchanged even when the aqueous phase is highly alkaline. The enhanced extraction efficiency appears to be related to the increased lipophilicity of the complexing agent. The distribution ratios for La-3+, E u ~ + and L u ~ + in the synthetic,INALYST. APRIL 1989. VOL. 113 453 sea water were calculated to be 1.5 x 102, 2.9 x 10' and 3.7 X 102, respectively (Table 1). The distribution ratio for La'+ in this instance is large enough such that over 94"/0 of La?+ in the sea water can be extracted into the organic phase at an aqueous to organic ratio of 10 : 1.The dependence of the extraction of Lu;+ by I1 on pH is shown in Fig. 2. I n comparison with I , the extraction plot shifts slightly to lower PIT. The lanthanide complexes extracted into the organic phase by I1 can be back-extracted easily into an aqueous phase at a pH level of less than 3. The rates of both the extraction and back-extraction processes are fast, requir- ing only cu. 1 min of vigorous shaking virtually to complete the extraction. In a real operation, the mixture is normally placed in the mechanical shaker for about 2 min for either of the extraction processes. According to our experiments, the volume of the organic phase to the acid solution during the b ac k - e x t r a c t i on process s h o u 1 d n c) t ex ce e d 1 0.o t h e rw i se incomplete recovery of the lanthanides can result. Therefore, the back-extraction process is capable of providing a 10-fold increase in the lanthanide concentration. A similar concentra- tion factor of 10 can also be obtained from the first step of the extraction. Using the two-step extraction procedure with I1 as the chelating agent, a total pre-concentration factor of 100 can be achieved for the lanthanides. It is notekvorthy that the sodium present in the sea water was only extracted in very small amounts (p.p.m. levels) and bromine was not detectable in the back-extracted acid solution. Based on the activity of 'JNa observed in the sea water experiments, we estimated that the relative extraction efficiencies of Na+iLu'+ under our experimental conditions were < 1 0 - 3 .This is important for analytical applications, because sodium and bromine are two of the major interfering matrix species for NAA of natural waters. Extraction with 2-(sym-Dibenzo-16-crown-5-oxy)stearic Acid The efficiency of extraction of La-'+ from the synthetic sea water with 111 in chloroform was better than 97% with an aqueous to organic phase ratio of 10 : 1. The distribution ratios for La'+, Eui+ and Lu-i+ were calculated to be 2.9 x 102. 3.8 X 102 and 1 .0 x 103, respectively (Table 1). The preference of extracting Lu'+ over La3+ was still observed in 111. The large C1(,H3; group attached to the side-arm should make the lanthanide complexes more lipophilic which is the likely cause of the enhancement of the distribution ratios.The lanthanide complexes in the organic phase can also be back-extracted into HNO; solution at pH < 3, suggesting that the exchange reaction between the complexed form and proton is not significantly affected by the large alkyl group in the side-arm. The dependence of the extraction of Lu-'+ by 111 on pH is shown in Fig. 2. The plot shifts further toward the lower pH levels relative to those of I and 11. Conclusions Increasing the length of the side-arm alkyl group in sym- dibenzo- 16-crown-S-oxyacetic acid increases the lipophilicity of the crown ether carboxylic acid and enhances the distribu- Table 1. Distribution ratios ( U ) of lanthanides with different crown ether carboxylic acids U Macrocyclic pol ycther La Eu I, u I .. . . . . . . 5.0 1 0 x 10' 2.0 x 10' I1 . . . . . . . . 1.5 x 10' 2.9 x 10' 3.7x 10' I11 . . . . . . . 2.9 x 10' 3.8 x 102 1.0 x 10' tion ratio of the lanthanide - macrocycle complexe5 in the organic phase. Preferential complexation with heavier lan- thanides was observed in this system. The macrocycles also show extremely low affinities for alkali metals. The CJ- and C,(,-substituted crown ether carboxylic acids, with high extraction efficiency and lipophihcity, are potential extraction agents for selective concentration of lanthanides from natural water systems. Since the extraction is reversible, lanthanides extracted into the organic phase can be back-extracted into an acid solution, thus providing a large pre-concentration factor (>loo) for chemical analysis. It is calculated that the two-step extraction process described in this paper combined with NAA can detect 10-1-10-3 pg of lanthanides in 200 ml of natural waters depending on the intrinsic sensitivities of different lanthanides to neutron activation. This research was supported in part by a grant frotn the Idaho State Board of Education. Neutron irradiations were per- formed at the Washington State University Nuclear Radiation Center under a Reactor Sharing Program supported by the Department of Energy. 1. 2. 3. 4. 5 . 6 7. 8 9. References Strzelbicki, J . , and Bartsch. R. A . , Anal. Chem., 1981, 53, 1894. Strzelbicki, J . , and Bartsch, K. A , , And. Chrm., 1981, 53, 2247. Manchanda, V. K., and Chang. C. A . , Anal. Chem., 1986,58, 2269. Tang. J . , and Wai, C. M., And. Chem., 1986, 58, 3233. Haskin, L. A.. Wildeman. T. R., and Haskin, M. A , , J . Radiut. Chem., 1968, 1, 337. Bartsch, R. A.. Heo, G. S . , Kang, S . I . , Liu, Y., and Strzelbicki. J . , J . (Irg. Chem., 1982, 47, 457. "Water Analy5ts by Atomic Absorption," Varian Techtron. Palo Alto. CA, 1972. Mok, W. M., Shah. N. K.. and Wai, C. M.,Anal. Chem., 1986, 58, 110. Strzelbicki. J . , and Bartsch, R. A., Anal. ('hem., 1981, 53, 225 1, Paper 8/034040 Received August 23rd, 1988 Accepted November l l t h , 1988