ANALYST, JANUARY 1993, VOL. 118 85 Column Preconcentration of Cobalt in Alloys and Pepperbush Using 2-(5-Bromo-2-pyridylazo)-5-diethylaminophenol and Ammonium Tetraphenylborate Adsorbent Supported-on Naphthalene With Subsequent Determination Using Atomic Absorption Spectrometry Masatada Satake Faculty of Engineering, Fukui University, Fukui 910, Japan Tohru Nagahiro Department of Chemistry, Himeji Institute of Technolog y, Himeji 671-22, Japan Bal Krishan Puri Department of Chemistry, Indian Institute of Technology, Hauz Khas, New Delhi- 11007, India A method has been established for column preconcentration and determination of cobalt using an ion-pair produced from an ammonium cation and a tetraphenylborate (TPB) anion supported on naphthalene in a simple funnel-tipped glass tube.Cobalt forms a water-soh ble chelate cation with 2-(5-bromo-2-pyridylazo)-5- diethylaminophenol (5-Br-PADAP). The chelate cation is retained as a water-insolu ble Co-5-Br-PADAP-TPB complex on the surface of the naphthalene, which is packed in a column. Cobalt is quantitatively retained on the ammonium tetraphenylborate adsorbent supported on the naphthalene in the pH range 3.0-8.0 at a flow rate of 1 ml min-1. The solid mass is stripped from the column with 5 ml of dimethylformamide (DMF) and the cobalt determined by flame atomic absorption spectrometry at 240.7 nm. The calibration graph is linear over the concentration range from 1 to 20 pg of cobalt in 5 ml of final DMF solution. Seven replicate determinations of 12 pg of cobalt gave a mean absorbance of 0.124 with a relative standard deviation of 1 .O%.The sensitivity for a 1 % absorption is 0.085 pg ml-1. The effect of foreign ions was studied and the proposed method applied to the determination of cobalt in certified reference material samples of alloy, steel and pepperbush. Keywords: Cobalt determination; ammonium tetraphen ylborate adsorbent; alloy, steel and pepperbush; nap h t halen e column p reconcen tra ti0 n ; 2- (5- b rom 0-2-p yrid ylazo)-5-dieth yla mino p hen o I; flame a to mic absorption spectrometry Sodium tetraphenylborate (TPB) or its derivative has been utilized in the gravimetric determination of some alkali and univalent metal ions.’-3 Recently it has been utilized as the counter ion in the extraction and adsorption of iron, cobalt, copper complexes of 2,2’-dipyridyl, 1,lO-phenanthroline (1,lO-phen) , 3-(4-phenyl-2-pyridyl)-5,6-diphenyl- 1,2,4-tri- azine, 2,6-bis(2-pyridyl)-4-(4-methoxyphenyl)pyridine, 2,4,6- tris(2-pyridyl)-1,3,5-triazine and 3-(2-pyridyl)-5,6-diphenyl- 1,2,4-triazine (PDT) into molten naphthalene4-9 or on microcrystalline naphthalenel(L15 as these ternary complexes cannot be extracted into the usual organic solvents owing to their poor solubility.The various advantages of these tech- niques have been discussed previously .4-15 A survey of the literature reveals that column methods have been reported for the separation and preconcentration of metal ions using various adsorbents such as activated car- bon,16 green tea leaves,l7 a chelating resin,’* thiol cotton19 and polythioether foam.”) Although some of these methods are fairly effective, once prepared they are reusable, unlike the material used in this procedure.Furthermore, introduc- tion of Chelex-100” and an ion-exchange resin22 minicolumn into the flow injection system increases the sensitivity to between ppb and ppt level, thus allowing natural waters to be analysed. Column methods have already been reported for the preconcentration of trace metals using various adsorbents, e.g., 2-mercaptobenzothiazole,*~ 2,4,6-tris(2-pyridyl)-1,3,5- triazine tetraphenylborate,24 mixed ligands of dimethylgly- oxime and acenaphthenequinone dioxime ,25 benzyldimethyl- tetradecylammonium perchlorate26 and tetradecyldimethyl- benzylammonium-l,2-dihydroxybenzene-3,5-disulfonate .27 In this paper, a highly selective and sensitive preconcentra- tion method has been developed, that uses 2-(5-bromo-2- pyridylazo)-5-diethylaminophenol (5-Br-PADAP) as the complexing agent and in addition to the (NH4+) (TPB-) ion-pair supported on naphthalene.Numerous reagents have been used for the spectrophotometric determination of cobalt. Among them, 5-Br-PADAP and its derivatives are highly selective for cobalt.28 Once the cobalt complex cation is formed, it does not decompose even in strongly acidic medium (as opposed to other metal complexes), this results in a high selectivity for cobalt. The 5-Br-PADAP reacts with cobalt to form a water-soluble complex cation, but in the presence of the TPB anion it forms a water-insoluble complex (Co-5-Br- PADAP-TPB). Preliminary observations revealed that the Co-5-Br-PADAP complex cation could be quantitatively retained on an ammonium tetraphenylborate adsorbent sup- ported on naphthalene.The method developed is very simple as only the cationic cobalt complex of 5-Br-PADAP is passed through the naphthalene adsorbent, which is packed in a funnel-tipped glass column. The metal complex along with the naphthalenc is dissolved out from the column with a small volume of solvent [2-5 ml of dimethylformamide (DMF)] and can be directly aspirated into the flame of an atomic absorption spectrometer. The operating parameters have been evaluated and the proposed method applied to the determination of cobalt in various certified reference material samples. It can also be employed for various other biological and environmental samples.Experimental Apparatus A Perkin-Elmer Model 403 atomic absorption spectrometer and a Toa-Dempa HM-SA pH meter were used. All absorp- tion measurements were performed under the following qperating conditions: wavelength, 240.7 nm; slit setting, 3(7 A); current, 10 mA; acetylene flow setting, 15 (pressure, 0.386 ANALYST, JANUARY 1993, VOL. 118 kg cm-2); and air flow setting, 55 (pressure, 2.1 kg cm-2). A funnel-tipped glass tube (SO X 6 mm i.d.) was used as the column, which was plugged with poly(propy1ene) fibres and then slurry-packed with the naphthalene material. The column loaded with the naphthalene was lightly compressed with the flat end of a glass rod so that its height was 1 .O-1.2 em. Reagents All the reagents were of analytical-reagent grade.Standard cobalt solution (3 ppm) was prepared by diluting 1000 ppm atomic absorption standard cobalt chloride solution with doubly distilled water. A 0.01% solution of 5-Br-PADAP in ethanol was prepared. Buffer solutions of pH 3-6, 6-8 and 8-11 were prepared by mixing appropriate ratios of a 0.5 rnol 1-1 acetic acid and 0.5 mol 1-1 ammonium acetate solution, 0.1 mol 1-1 sodium dihydrogen phosphate solution and 0.1 rnol 1- dipotassium hydrogen phosphate solution and 0.5 mol 1-1 ammonia solution and 0.5 mol 1-1 ammonium acetate solution. Doubly distilled water was used throughout. Preparation of Naphthalene-NH4-TPB Adsorbent A solution of naphthalene was prepared by dissolving 20 g of naphthalene in 40 ml of acetone on a hot-plate stirrer at 35 "C.This solution was transferred into 1500 ml of distilled water, containing 100 ml of 1 mol 1-1 ammonium acetate-1 rnol 1-1 ammonia solution (pH 9 . 9 , in a fast stream with continuous stirring at room temperature, and to it was added 500 ml of an aqueous solution of 1.7 g of TPB in the same manner. The naphthalene material coprecipitated with NH4+ and TPB-, the mixture was stirred for about 2 h and then allowed to stand for 2 h. The supernatant solution was decanted off and the naphthalene washed twice with distilled water. The slurry of naphthalene in water was stored in a bottle until required. General Procedure An aliquot of solution containing 1-20 pg of cobalt was made up to 13 ml with distilled water in a 20 ml beaker to which were added 1.0 ml of the 0.01% 5-Br-PADAP in ethanol solution and 0.5 ml of 0.1 moll-' phosphate buffer (pH 6.0-6.5).This solution was passed through a column loaded with naph- thalene-NH4-TPB at a flow rate of 1 ml min-1. The packing was washed with a small volume of distilled water and then aspirated strongly for 5 min, pushing down the naphthalene material with a flat glass rod to eliminate the excess water attached to the naphthalene. The metal complex along with the naphthalene was dissolved out from the column with 5 ml of DMF. The solution was aspirated into an air-acetylene flame and the absorbance measured at 240.7 nm against a reagent blank. Results and Discussion Retention Characteristics of NH4-TPB Sodium tetraphenylborate is soluble in water, it forms water-insoluble precipitates with some alkali metal ions such as K+, Rbi , Cs+ (but not Li+ and Na+), and univalent metal ions such as Ag+, TI+ and Cu+, but does not form precipitates with multivalent metal ions.It has been used as a gravimetric and volumetric reagent. Furthermore, it also reacts with onium salts such as NH4+ and its derivatives, e.g., ammines, quaternary ammonium salts, alkaloids and onium compounds, to form water-insoluble precipitates. The TPB- forms a weakly bonded ion-pair with NH4+ in aqueous solution and coprecipitates with microcrystalline naphthalene, as follows: From the experimental observation, the NH4-TPB ion- pair, produced from TPB and ammonium acetate in aqueous solution, when supported on naphthalene was unstable and NH4+ + TPB- $ (NHd+) (TPH-) (s).partly desorbed from the surface of the naphthalene in the column on passage of the buffer. Thus the NH,-TPB ion-pair was prepared in acetate buffer of pH 9.5. The adsorbent shows excellent adsorption characteristics for various metal complex cations such as Fe( 1 ,10-phen)32+ or CU(PDT)~+. In this work, TPB- is selected as the counter ion because of its high purity and moderate price. Reaction Conditions Keeping the other variables constant, experiments were carried out using 12 pg of cobalt. The adsorption of cobalt starts at pH 1.2 and is constant and maximum over the pH range 3.0-8.0; above pH 8.0, the adsorption is decreased. Addition of 0.3-2.0 ml of the phosphate buffer did not affect the adsorption of cobalt and use of 0.5 ml at pH 6.5 is recommended Various amounts of 0.01Y0 5-Br-PADAP in ethanol were added to the solution containing 12 pg of cobalt.Cobalt was quantitatively adsorbed onto the naphthalene over the con- centration range of 0.5-2.0 ml of reagent solution. Thus 1 .0 ml was used in all subsequent experiments. The flow rate was varied from 0.5 to 8 ml min-1. The adsorption of cobalt was not affected over this range. A flow rate of 1 mi min-1 was recommended. The effect of the volume of the aqueous phase on adsorption of cobalt was investigated by the general procedure. The adsorption was constant and maximum when the volume of the aqueous phase did not exceed 700 ml. In this case, cobalt is preconcentrated by about 140-fold for 5 ml of DMF solution. In the subsequent work, 15 ml of the aqueous phase were used for convenience.Choice of Solvent An attempt was made to dissolve out the naphthalene-Co-5- Br-PADAP-TPB complex from the column. As the solid mass was dissolved out with a small volume ( 5 mi) o f solvent it was essential to select a solvent in which the chelate was highly soluble and could be determined sensitively. The solid material was soluble in methanol, ethanol, butan-1-01, aceto- nitrile, dimethyl sulfoxide and DMF. The DMF was preferred because of the high solubility and sensitivity attained. Methanol and ethanol are better solvents, but the flame on the burner-head was unstable owing to their volatility. It was found that 2-3 mi of DMF were sufficient to dissolve the mixture, it also enhanced the sensitivity of the method by a further 2-fold.The surplus water attached to the naphthalene caused the absorbance to decrease by 10'3'0 and created an error in the determination, it was, therefore, necessary to eliminate the water by aspirating. Linearity, Sensitivity and Precision Based on the optimum conditions described above, the calibration graph was linear over the concentration range 1-20 pg of cobalt in 5 ml of DMF solution. Seven replicate determinations of 12 pg of cobalt gave a mean absorbance of 0.124 with a relative standard deviation of 1.0%. The sensitivity for 1% absorption was 0.085 pg ml-l (0.24 pg ml-1 for direct flame atomic absorption spectrometric measure- ment on the aqueous solution). Effect of Diverse Ions Sample solutions containing 12 pg of cobalt and alkali salts or metal ions were prepared individually and the General Procedure applied.The tolerance limit was set as the amount of diverse ion required to cause +3% error in the determina- tion of cobalt. The results obtained are given in Table 1. Of the salts examined, most could be tolerated up to gram or milligram levels. Disodium ethylenediaminetetraacetic acid (EDTA) (10 pg) could be tolerated. Of the metal ions studied,ANALYST. JANUARY 1993, VOL. 118 87 most metal ions could be tolerated up to a level of 1 mg; 100 mg of Ag+, Pt4+ and V-5+ could be tolerated. Copper was masked by the addition of thiourea. Thus, the proposed method is selective and can be applied to the determination of cobalt in various certified reference material (CRM) samples without any preliminary separation and can be employed for other complex materials.Determination of Cobalt in CRM Samples of Aluminium Alloy, Stainless Steels and Pepperbush The proposed method has been applied to the determination of cobalt in Nippon Keikinzoku Kogyo (NKK) CRM 920 Aluminium Alloy, Japanese Standards of Iron and Steel (JSS) CRM 651-7 Stainless Steel and National Institute for Environ- mental Studies (NIES) CRM No. 1 Pepperbush. A 0.054.1 g Table 1 Effect of diverse salt and ions Salt or ion CH3COONa.3H20, NaC1O4.Hz0, KN03 NaCl KI, thiourea Na2S04 NH4CI, KSCN, Na, K-tartrate Sodium citrate L-Ascorbic acid KCN Disodium EDTA Call, Mg", Al"', Mo"', Mn", Fe"' Pb". Bill', Cd", W"', Crvl, Ru"l Zn" . Cr"l , Pd" Ag', Ptlv, Vv Cu" KH2PO4 NHJF. GEDTAt K?C?OI. H20 Tolerance limit 1 g* Ig 500 mg* 500 mg 100 mg 10 mg* 10 mg 5 mg" 5 mg 3 mg 500 P8 10 I-18 500 I % 10 pg, 500 pgj 5 mg* 1 mg" 100 * Maximum value tested: Co, 12 pg, pH, 6.5; and 0.01% t GEDTA = glycol ether diaminetetraacetic acid.$ Masked with 500 mg of thiourea. 5-Br-PADAP, 1 .O ml. sample of CRM 920 and CRM 651-7 was completely dissolved in a minimum volume (about 8 ml) of hydrochloric acid (1 + 1) by heating on a water-bath. To this was added 1 ml of 30% hydrogen peroxide. The excess hydrogen peroxide was decomposed by heating the solution on a water-bath. The solution was cooled, filtered if necessary and diluted to 100 ml in a calibrated flask. A 2 g amount of NIES CRM No. 1 Pepperbush was completely dissolved by heating in a minimum volume of concentrated nitric acid (about 25 ml) and 1 ml of concen- trated perchloric acid.The solution was evaporated to very small volume. To this a small volume of water was added. The solution was cooled, filtered and diluted to 50 ml in a calibrated flask. An aliquot (1-2 ml) of each sample was transferred into a 20 ml beaker, and to this were added: suitable amounts of masking agents, 1.0 ml of 0.01% 5-Br-PADAP in ethanol solution and 0.5 ml of the buffer. After being left to stand for 20-30 min, the sample solution was adjusted to pH 1.0 with concentrated nitric acid in order to decompose any interfering metal complexes and then the General Procedure was applied. The results are given in Table 2. These results are in agreement with the certified values. Conclusion A solid ion-pair compound produced from NH4+ and TPB- on naphthalene provides a simple and economical method for column preconcentration of cobalt in alloys and biological samples.The proposed method is the first attempt to separate and concentrate cobalt using the rcaction of Co-5-Hr-PADAP complex cation with NH4-TPB supported on naphthalene. As the proposed method requires only simple glassware, such as a funnel-tipped glass tube and small volume beakers, and as only a small volume of organic solvent is used for the dissolution of the complex, it is vcry economical. The sensitivity and selectivity of the method might be further improved by using alternative optical analytical techniques of analysis such as flameless atomic absorption spectrometry. Table 2 Analysis of samples for cobalt Concentration of cobalt (%) Sample Stainless Steel JSS 651-7 NKK920 Al u m i n i u in Alloy NIESNo.1 Pepperbush11 Composition (%) C, 0.047; Si, 0.72; Mn. 1.72; P, 0.028; Cr, 18.60; S, 0.0063; Mo, 0.84; Cu, 0.082; Al, 0.002; N , 0.0312; and Ni, 9.20 Si, 0.78: Fe, 0.72; Cu, 0.71; Mn, 0.20; Mg, 0.46; Cr, 0.27; Zn, 0.80; Ti, 0 . 15; Sn, 0.20; Pb. 0.10; V. 0.15; Sb, 0.01; Bi. 0.06; Ga, 0.05; Ni, 0.29; Sb, 0.01; and Ca, 0.03; K, 1 .S I -t 0.06; Mg 0.4O8 t 0.020; Ca, 1.38 k 0.07; and Mn, 0.203 k 0.017. Fe, 205 f 17; Zn, 340 t 20; Ba, 165 f 10; Na, 106 2 13; Rb, 75 f 4; Sr, 36 t 4; Co, 23 t 3; Cu, 12 2 1; Ni, 8.7 t 0.6; Cd, 6.7 t 0.5; Pb, 5.5 f 0.8; As, 2.3 f 0.3; P, (1100);Cr.(1.3);Cs. (1.2); TI. (0.13); and Hg, (0.056) v g g - * Certified value Found* 0.22 0.240 t 0.003-t 0.10 0.099 k 0.003$ 23 k 3 pgg-1 25 -t 29 pgg-1 * Mean of four determinations t 100 mg o f thiourea and 1 0 mg of ammonium citrate were added as masking agents at pH 4.04.5.$ 2 ml of 20% triethanolamine solution were added as masking agents at pH 8.0-8.2. 10 mg of ammonium citratc, 2.0 ml of 0.1 mol I- sodium pyrophosphate solution and 1 .O ml of 0.01 mol I-' GEDTA solution were added as masking agents at pH 4.0-4.5. Before passing the sample solution to the column, the sample solution was readjusted to pH 1 with conccntrated hydrochloric acid in order to decompow interfering metal complexes. 11 Values in parentheses are tor reference and are not certified.88 ANALYST, JANUARY 1993, VOL. 118 1 2 3 4 5 6 7 8 9 10 11 12 13 14 References Barnard. Jr., A.J.. Chemist-Analyst, 1955,44, 104. Barnard, Jr., A. J., Chemist-Analyst, 1958, 47, 46. Advances in Analytical Chemistry and Znstrumentation, ed, Reilley, C. No, Interscience, New York. 1960, vol. 1. Satake, M., Nagahiro, T., and Puri, B. K., Analyst, 1984, 109, 31. Chang, L. F., Satakc, M., Kuwamoto, T., and Puri, B. K., Microchem. J . , 1986. 33, 46. Nagahiro, T., Uesugi. K., Satake, M., and Puri, B. K., Bull. Chem. Soc. Jpn., 1985, 85, 1115. Puri, B. K., Mehra, M. C., and Satake, M., Chem. Anal. (Warsaw), 1986, 31, 185. Mehra, M. C., Satake. M., Chang, L. F., and Kuwamoto. T., Acta Cient. Venez., 1984,35, 232. Nagahiro, T., Satake. M., and Puri, B. K., Zndian J . Chem., 1986, 25A. 99. Satakc, M., Mehra, M. C., Singh, H. B., and Fujinaga, T., Bunseki Kagaku, 1983, 32, E165. Chang, L. F., Satake, M., Puri, B. K., and Bag, S. P., Bull. Chem. Soc. Jpn.. 1983, 56, 200. Lin, J. L., Chang. L. F., Katyal, M., and Satake, M., Fresenius’ Z. Anal. Chem., 1984,319, 308. Nagahiro, T., Satakc, M., Lin, J . L., and Puri, B. K., Analyst, 1984, 109, 163. Lin, J . L., Satake, M., and Puri, B. K., Analyst, 1985,110,1351. 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Mehra, M. C., Nagahiro, T., and Satake, M., Microchem. J., 1986, 33, 198. Smith, S., Nelissen, J . , and van Grieken, R. E., Anal. Chim. Acta, 1979, 111, 215. Kimura, M., Yamashita, H . , and Komada, J.. Bunseki Kaguku. 1986, 35, 400. Florence, T. M., and Batley, G. E., Talanta, 1976, 23, 179. Yu, Mu-Qing. and Liu, Gui-Qin, Talanta, 1983. 30, 265. Khan, A. S., and Chow. A.. Talanta, 1986, 33, 182. Olsen, S., Pesscnda, L. C. R., RiiiCka, J . , and Hansen, E. H., Analyst, 1983, 106, 905. Kamson, 0. F., and Townshend, A., Anal. Chim. Acta, 1983, 155, 253. Satake, M., Ishida, K., Puri, B. K., and Usami. S., Anal. Chem., 1986,58, 2502. Usami, S., Yamada, S., Puri, B. K., and Satake, M., Mikro- chim. Acta, Part I , 1989, 263. Usami, S., Fukami, T., Kinosita, E., Puri, B. K., and Satake, M., Anal. Chim. Acta, 1990, 230, 17. Miura, J., Arima, S., and Satake, M., Analyst, 1990,115, 1191. Satake, M., Nagahiro, T., and Puri, B. K., Tulanta, 1992, 39, 1349. Taguchi, S., and Goto, K., Tulanta, 1980, 27,819. Paper 2f02284B Received May I , 1992 Accepted October 2, 1992