ANALYST, APRIL 1988, VOL. 113 581 Separation and Concentration of Beryllium by Coprecipitation with Hafnium Hydroxide Prior to Determination bv GraDhite Furnace Atomic Absorption Spectrometry Joichi Ueda and Toshio Kitadani Department of Chemistry, Faculty of Education, Kanazawa University, Marunouchi, Kanazawa 920, Japan Hafnium hydroxide coprecipitates 0.01-0.2 pg of beryllium from 100-400 mi of sample solution at pH 6.&10.5. The presence of hafnium quantitatively increases the atomic absorbance of beryllium by about 1.5 times, and the reproducibility of the measurement is improved. The calibration graph is linear for 0.4-8 ng mi-1 of beryllium. The interference induced by large amounts of aluminium can be eliminated by raising the pH of the solution to 13-13.5 with sodium hydroxide solution after the coprecipitation procedure has been carried out at about pH 9.5.Several other ions tested did not produce serious interference effects. This method is applicable to the determination of trace amounts of beryllium in water samples and in aluminium. Keywords: Beryllium determination; graphite furnace atomic absorption spectrometry; coprecipitation; hafnium hydroxide a Coprecipitation is a useful way of concentrating trace ele- ments and a variety of coprecipitants have been proposed.1 We have been examining the use of hafnium hydroxide as a coprecipitant and have suggested that it is a suitable collector for lead, cadmium and copper.*-4 We have found that hafnium hydroxide is also an excellent collector of trace amounts of beryllium and that the coprecipitated beryllium can be determined satisfactorily by graphite furnace atomic absorp- tion spectrometry.This method is simple, and the sensitivity and the reproducibility are improved by the presence of the coprecipitant. Several coprecipitants have been proposed for concentrating beryllium,s11 and zirconium hydroxidelo has been used prior to determination by graphite furnace atomic absorption spectrometry. However, the use of zirconium hydroxide is disadvantageous because of its poor solubility in acids. Hafnium hydroxide dissolves easily in acids and can collect smaller amounts of beryllium in water than zirconium hydroxide. It is also possible to utilise this method for the determination of trace amounts of beryllium in large amounts of aluminium. This paper describes the fundamental con- ditions for the coprecipitation of beryllium with hafnium hydroxide and for the graphite furnace atomic absorption spectrometric determination of beryllium.Experimental Apparatus A Hitachi 170-70 Zeeman effect atomic absorption spec- trometer with a Hitachi beryllium hollow-cathode lamp was used for the atomic absorption measurements, and a Hitachi- Horiba Model M-5 glass electrode pH meter for the pH measurements. Reagents All chemicals were of anal ytical-reagent grade. Standard beryllium solution. A solution containing 1 mg ml-1 of beryllium was prepared by dissolving metallic beryllium (Wako Pure Chemicals) in a small amount of hot sulphuric acid and diluting with distilled water. This solution was further diluted to the desired concentration with small amounts of nitric acid and distilled water. Hafnium solution.A solution containing about 5 mg ml-1 of hafnium was prepared by dissolving hafnium chloride (Naka- rai Chemicals) in distilled water and standardised by com- plexometric back-titration with a standard thorium solution using xylenol orange as indicator. Recommended Procedure To a sample solution (100-400 ml) containing 0.01-0.2 pg of beryllium, 20 mg of hafnium are added and the pH of the solution is adjusted to about 9.5 with aqueous ammonia (1 + 1). After the precipitate has settled, the solution is filtered through a 3G4 sintered-glass filter. The precipitate is washed with a small amount of aqueous ammonia (pH 9.5) and dissolved in 1 ml of concentrated nitric acid.The solution is made up to 25 ml with distilled water and the atomic absorbance of beryllium is measured under the operating conditions shown in Table 1. The blank is also taken through the same procedure, using distilled water instead of a sample solution. Results and Discussion Optimisation of Instrumental Parameters The optimum conditions for the measurement of the atomic absorbance of beryllium were examined using a 0.6 M nitric acid solution containing 8 ng ml-1 of beryllium and 0.8 mg ml-1 of hafnium. The current and the drying time, which gave almost constant absorbance, were 22-25 A and 4&80 s, respectively. At the charring stage, the peak height of beryllium increased with an increase in the heating current, reached a maximum at 110-140 A (Fig.1) and then remained almost constant with 20-70 s of heating time. Fig. 1 shows the results obtained in the presence and absence of hafnium. It is recognised that the optimum charring current can be raised to 140 A without any loss of beryllium in the presence of hafnium. At the atomisation stage, the maximum peak height was obtained at Table 1. Operating conditions for the atomic absorption spectrometer Analytical wavelength Lampcurrent . . Slitwidth . . , . Argon flow-rate Sheathgas . . Carriergas . . Injectionvolume . . Cuvette . . . . Dryingconditions . . Charring conditions Atomising conditions .. .. .. .. .. .. . . * . . . .. 234.9 nm 15 mA No. 3 (2.2 nm) 3 1 min-1 0 ml min-1 Uncoated tube type graphite furnace 24 A (cu. 170 "C), 45 s llOA(ca.1180°C),30s 310 A (ca. 2830 "C), 5 s 10 pl582 ANALYST, APRIL 1988, VOL. 113 r O fi 40 80 120 160 Charring current/A Fig. 1. Effect of charring current on beryllium peak height examined with a 25-ml solution containing 0.2 pg of beryllium, 1 ml of concentrated nitric acid and (A) 20 or (B) 0 mg of hafnium. Atomising conditions: 310 A, 5 s I 280 310 Atomising currentlA Fig. 2. Effect of atomising current on beryllium peak height examined with a 25-ml solution containin 0 2 pg of beryllium, 1 ml of concentrated nitric acid and (A) 20 or (By 0 mg of hafnium. Charring conditions: (A) 110 A, 30 s and (B) 60 A, 30 s 0 20 40 Hf/mg per 25 ml Fig. 3. Effect of amount of hafnium on beryllium eak height examined with a 25-ml solution containing 0.2 p of berylEum, 1 ml of concentrated nitric acid and various amounts of hafnium 70 k I 0 2 4 Volume of acid/ml Fig.4. Effect of acids on beryllium peak hei ht examined with a 25-ml solution containing 0.2 pg of beryllium, 28 mg of hafnium and (A) concentrated nitric acid or (B) concentrated hydrochloric acid Table 2. Effect of foreign ions on the determination of 0.2 pg of beryllium in about 100 ml of water Amount Beryllium Amount Ion addedrng found*/pg Ion added/ m g Ll+ . . . . 1 Na+ . . . . 1200 K+ . . . . 520 Mg2+ . . . . 200 Ca2+ . . . . 250 Sr2+ . . . . 1 Ba2+ . , , . 1 AP+ . . . . 1 AP+ . . . . 0.5 Ga3+ . . . . 1 In3+ . . . . 1 Sn4+ . . . . 1 Pbz+ . . . . 1 Sb3+ . . . . 1 0.198 f 0.002 0.193 k 0.002 0.200 f 0.005 0.197 f 0.003 0.203 f 0.004 0.199 k 0.001 0.210 k 0.003 0.220 f 0.005 0.199 f 0.004 0.200 f 0.004 0.196 f 0.004 0.208 f 0.003 0.208 f 0.004 0.206 f 0.003 * Mean f standard deviation of triplicate determinations.Bi3+ . . . . cuz+ * . . . Zn2+ . . . Cd2+ . . . . La'+ . . . . zrj+ . . . . Th4+ . . . . Cr3+ . . . . MoV1 . . . . WV' . . . . Mn2+ . . . . Fe3+ . . . . coz+ . . . . Ni*+ . . . . 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Beryllium found*/pg 0.206 f 0.005 0.206 f 0.003 0.200 f 0.003 0.203 k 0.005 0.201 k 0.002 0.203 f 0.003 0.207 f 0.003 0.204 f 0.002 0.197 f 0.005 0.200 f 0.006 0.201 f 0.003 0.199 f 0.003 0.202 f 0.004 0.200 f 0.002 310 A and from 3 to 10 s. A change of the sheath gas (Ar) flow-rate from 1 to 5 1 min-1 had no effect on the beryllium absorbance, but an increase in the carrier gas flow-rate up to 80 ml min-1 gradually decreased the absorbance.Of the three slit widths, no. 1 (0.4 nm), no. 2 (1.1 nm) and no. 3 (2.2 nm), slit no. 3 was selected because of smaller noise although the peak height decreased slightly. The optimum instrumental parameters are summarised in Table 1. Study of Optimum Conditions for Coprecipitation The necessary amount of hafnium for the coprecipitation was examined with sample solutions (100400 ml) containing 0.01 01 0.2 pg of beryllium, according to the recommended procedure. It increased with increasing sample volume, and more than 10 mg of hafnium should be added for the quantitative collection of beryllium from 400 ml of sample solution. The presence of hafnium improved the repro- ducibility of the atomic absorption measurements of beryl- lium, and the relative standard deviation of the peak heights obtained from ten repeated measurements for 8 ng ml-1 of beryllium was 1.8%, compared with 3.8% obtained in the absence of hafnium.Further, the sensitivity of the beryllium determination was also increased by the presence of hafnium, i.e., the peak height was about 1.5 times higher than that in the absence of hafnium (Figs. 1 and 2). This peak height was little affected by various amounts of added hafnium from 0.1 to 2 mg ml-1 (Fig. 3). Hence, 20 mg of hafnium were used for the coprecipitation. To determine the optimum pH of the coprecipitation, the recovery of beryllium was studied in the pH range 4.0-11.8 with a solution (about 100 ml) containing 0.2 pg of beryllium. The beryllium was almost completely collected at pH 6.0- 10.5; hence, the pH was adjusted to about 9.5 with aqueous ammonia in further experiments.The effect of standing time on the recovery was slight, and recovery was almost 100%ANALYST, APRIL 1988, VOL. 113 583 Table 3. Recovery of beryllium from spiked water samples Be added (10 ng) Sample volume/ml Be found*/ng RSD,% Be found*/ng RSD, % Be added (200 ng) Sample Distilledwater . . . . 100 Distilled water . . . . 400 River water Kanakusaririver . . 100 400 Asanoriver . . . . 100 400 Seawater . , . . . . 100 Seawater . . . . . . 400 * Average of five replicate determinations. 10.1 9.8 10.6 9.7 10.0 9.6 9.8 9.5 5 8 6.0 6.7 6.2 5.6 5.6 5.5 5.5 200.0 3 0 203.0 3 0 198.8 2.5 203.2 3.5 205.1 2.8 201.9 2.7 199.5 2.7 195.2 3.4 Table 4. Recovery of beryllium from large amounts of aluminium.Coprecipitation was carried out with about 60 ml of sample solution Amount of aluminiumhg 1 10 10 30 50 50 50 Beryllium added/ng 100 10 200 10 10 100 200 Beryllium found*/ng 99.0 10.3 198.0 9.8 10.2 97.4 203.3 * Average of five replicate determinations. RSD , % 3.3 6.1 1.6 6.8 8.2 2.4 2.1 from a few minutes to 10 h of standing. The precipitate was filtered off after it had settled. Hafnium hydroxide was dissolved satisfactorily in both nitric and hydrochloric acid3 (Fig. 4). In these experiments, 1 ml of concentrated nitric acid was used. Calibration Graph A straight line passing through the origin was obtained over the concentration range 0.4-8 ng ml-1 of beryllium using the suggested method. The relative standard deviation was 3.0% for 0.2 pg of beryllium in about 100 ml of sample solution (five observations) and the detection limit (signal to noise ratio of two) was 6.25 pg ml-1 of beryllium in 400 ml of initial sample solution.Interferences The effect of 27 foreign ions on the determination of 0.2 pg of beryllium was examined by the recommended procedure (Table 2). Large amounts of sodium, potassium, magnesium and calcium and of up to 0.5 mg of aluminium did not interfere with the determination. No other ions tested produced serious interference effects, even at a concentration 5000 times that of beryllium. Recovery of Beryllium from Spiked Water Samples The utility of this method was evaluated by investigating the recovery of beryllium by the recommended procedure from river and sea water samples spiked with beryllium.The samples were filtered through a Toyo Roshi TM-2p membrane filter (pore size 0.45 pm) as soon as possible after sampling, and acidified with hydrochloric acid to about pH 2 for storage. The results obtained are shown in Table 3, and it seems that the proposed method is applicable to the analysis of these water samples. Recovery of Beryllium in Aluminium The presence of up to 0.5 mg of aluminium does not interfere with the determination of beryllium (Table 2), but amounts greater than 0.5 mg gave a positive error. However, it was found that the effect of large amounts of aluminium can be eliminated by adjusting the pH of the sample solution. The procedure is as follows. Beryllium (0.01-0.2 pg) is copre- cipitated at about pH 9.5 according to the recommended procedure in a solution containing large amounts of alum- inium.After standing for at least 10 min, the pH of the solution is raised to 13-13.5 with sodium hydroxide solution to dissolve aluminium hydroxide and the solution is filtered through a 3G4 sintered-glass filter. The subsequent procedure is as given in the recommended procedure. The recovery was good (Table 4), and this method is applicable to the determination of trace amounts of beryllium in aluminium. Conclusions From this study, hafnium hydroxide is recommended as a coprecipitant for pre-concentration prior to the determination of beryllium in water samples by graphite furnace atomic absorption spectrometry, owing to its excellent collecting ability. The presence of hafnium increases the atomic ab- sorbance of beryllium and improves the reproducibility of the measurement. The detection limit (signal to noise ratio of two) of this method is 6.25 pg ml-1 of beryllium in 400 ml of the initial sample solution. This method is also applicable to the determination of trace amounts of beryllium in aluminium. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. References Mizuike, A., “Enrichment Techniques for Inorganic Trace Analysis,” Springer-Verlag, Berlin, Heidelberg, 1983, pp. Ueda, J., and Kita, C., Bull. Chem. SOC. Jpn., 1985,58, 1899. Ueda, J., and Yamazaki, N., Bull. Chem. SOC. Jpn., 1986,59, 1845. Ueda, J., and Yamazaki, N., Analyst, 1987, 112,283. Shigematsu, T., Tabushi, M., and Isojima, F., Bunseki Kagaku, 1962, 11,752. Novikov, A. I., Ruzankin, V. I., and Khamidov, B. O., Dokl. Akad. Nauk Tadzh. SSR, 1969, 12,22; Anal. Abstr., 1971,21, 29. Lebedinskaya, M. P., and Chuiko, V. T., Zh. Anal. Khirn., 1973,28,863. Sudhalatha, K . , Talanta, 1963,10, 934. Dunaeva, Yu. N., Trudy Sverdl. Skh. Inst., 1964, 11, 509; Anal. Abstr., 1966, 13, 1015. Sato, A., and Saitoh, N., Bunseki Kagaku, 1977,26747. Plotnikov, V. I., and Safonov, I. I., Zh Anal. Khzm., 1978,33, 2350. 61-66. Paper A71363 Received August 18th, 1987 Accepted November 2nd, 1987