and the inner wall of the separator was flushed with about 50 ml of water, which was again drained off. The cyclohexane layer was then run through a plug of cotton for spectrum taking in the range of 260-370 nm, using cyclohexane as reference. The sodium cyclamate content was determined by comparing absorbance at 314 nm with the standard curve. PREPARATION OF STANDARD CURVE. Volumes of 100 ml of water containing 2.00, 4.00, 8.00, 12.00, 16.00, and 20.00 mg of sodium cyclamate were carried through the above procedure for the assay of cyclamate. The amount (mg) of sodium cyclamate used was plotted against absorbance at 314 nm. A straight line obeying Beer's law was obtained. RESULTS AND DISCUSSION
One-hundred milliliter solutions of various foods with 5 and 10 mg of sodium cyclamate added, were tested. Results are in Table I. Slight interference was found with some foods. This is why the recovery of added cyclamate is sometimes a little higher than 100 %. Amino acids, urea, ammonia, various aliphatic amines, and food additives such as salicylic acid, benzoic acid, p-hydroxybenzoates, potassium sorbate, saccharin, and dulcin were tested for interference. Only dulcin was found to interfere slightly, but its presence was revealed by a brown coloration of the water layer in the step when cyclohexane layer was shaken with 50 ml of 0.5N NaOH. As is shown in Figure 1, N,N-dichlorocyclohexylamine has two absorption peaks. Although the absorptivity at 222 nm is much larger than at 314 nm, wavelength 314 nm was adopted because interference was found negligibly slight at 314 nm for the various foods we had tested. A spectrum with a peak at 314 nm and a valley at 274 nm definitely gives a qualitative picture, disregarding the slight interference. When 100 ml of ethyl acetate was used to extract 100 ml of water containing a certain amount of sodium cyclamate (1-20 mg), it was found 7.5 ml of concentrated sulfuric acid was the critical amount needed to acidify the aqueous solution so that the maximum proportion of cyclamic acid was extracted (about 46.5 %). Beyond this critical amount, when 10, 15, and 20 ml were tested, the maximum proportion of cyclamic acid being extracted remained the same. The reason that 10 ml of sulfuric acid was used in our procedure is that after possible consumption of part of the sulfuric acid by the sample solution, the remaining sulfuric acid would acidify
Table I. Average Recovery of Sodium Cyclamate, Obtained from Three Analyses for Each Cyclamate-Added Sample Average recovery, 5 mg added 10 mg added Sample (100 ml). Guava juice 105.0 0 . 2 103.5 i 0 . 4 Orange juice 104.5 3~ 0 . 5 103.0 i 0 . 2 Tomato juice 100.2 i 0.1 100.0 i 0 . 2 100.5 + 0 . 2 100.9 k 0 . 3 Mango juice Cola A 104.8 + 0 . 3 104.3 =k 0 . 1 Cola B 104.4 i 0 . 3 103.7 f 0 . 2 Soy sauce A 104.2 k 0 . 6 104.0 i 0 . 5 Soy sauce B 103.5 i 0 . 4 102.4 i 0 . 3 Pork lean extract 99.7 i 0 . 2 99.2 i 0.4 Fish extract 100.3 i 0 . 4 100.1 f 0 . 3 5 Milk powder 102.2 + 0.5 101.7 i 0 . 5 3 0 z Vanilla ice cream 100.6 f 0 . 3 100.2 i 0 . 2 Totalav 102.2 =t1 . 8 2 a Fruit juices and soy sauces were diluted by a ratio of 1 : 1.
*
the sample solution strongly enough to ensure maximumproportion extraction. Determination of cyclamate without previous extraction with ethyl acetate can be achieved only with some drinks. With others, as in the case of soy sauce, consumption of hypochlorite was too big to render the determination possible. Emulsion problems did usually occur with some food products. In extraction with ethyl acetate, suspensions of dairy products of considerably high concentration, led to serious emulsion. But clear ethyl acetate, although less in volume than originally used, was easily obtained by centrifuging followed by filtering the upper supernatant through cotton. For fruit juices and soy sauces, slight emulsion occurred in between the two layers, After draining off the water layer and the emulsion part, clear ethyl acetate was obtained by filtering through cotton. In case the clear ethyl acetate layer obtained is less than 80 ml, the last result of assay could still be obtained by correction to 80 ml. In the subsequent extractions with cyclohexane, emulsion did not occur. RECEIVED for review November 3, 1971. Accepted January 7,1972.
CORRESPONDENCE
I
A Simplified Separation of Strontium, Radium, and Lead from Environmental Media by Precipitation Followed by Fractional Elution SIR: Procedures for strontium-89 and -90 in media such as water, soil, food ash, and bone ash commonly take advantage of the insolubility of Sr(NO& in strong nitric acid to effect a separation of strontium from calcium. Sensitivity requirements in environmental analysis dictate amounts of sample often containing up to 3 grams of calcium. In addition, there is a restriction on the amount of strontium used as carrier, typically 20 mg, to minimize self-absorption of beta particles during counting. Thus the calcium :strontium ratio is often greater than 100 : 1. A suitable choice of nitric acid concentra-
tion and repeated treatments are needed to effect a satisfactory separation with minimum loss of strontium. An early study ( I ) established the optimum nitric acid concentration at about 80% wjw. This concentration is obtained by dilution of unpleasant, hazardous, fuming nitric acid (nominally 95 wjw). Since then, however, successful separations have been achieved using nitric acid at about the (1) H. H. Willard and E. W. Goodspeed, IND.ENG.CHEM., ANAL. ED., 8, 414 (1936).
ANALYTICAL CHEMISTRY, VOL. 44, NO. 12, OCTOBER 1972
2113
strength of the common laboratory reagent, 16N (70% wjw). Bruenger and coworkers (2), for instance, used the constant obtained by repeated evaporation boiling mixture 69 "03 and addition of the laboratory reagent during analysis. Pecently Baratta and Knowles (3) investigated the separation using nitric acid concentrations ranging from 65 to 80%. They recommend approximately 15N ( 6 7 z ) as the optimum concentration when using two separations, and they also confirm the use of 7 0 x "OB recommended in some procedures (4, 5). Their experimental data indicate that when 5 ml of H 2 0 , containing 1000 mg of Ca2+and 20 mg of Srzf, is adjusted to 70% "03, by addition of fuming nitric acid, strontium precipitates almost completely and more than 30 % (300 mg) of the calcium coprecipitates also. Therefore under these conditions more than 600 mg of Ca2+remain dissolved, and calculations indicate that if the volume of the 7 0 x H N 0 3 were increased to about 40 ml, then more than 1.5 grams of Ca could be held in solution with very little additional loss of Sr. Such calculations from Baratta and Knowles experimental data are in accord with our own experimental results, and the use of 70% H N 0 3 in the current routine procedures of this laboratory. In these procedures 20 mg of Sr2+is separated in high yield from prepared environmental samples containing up to 3 grams of Ca. The separation is accomplished in one step by digestion of the dried alkaline earth carbonate or phosphate precipitate with 40 ml of 70% HN03. An additional 40 ml of the acid is then used to transfer and wash the precipitate, thus reducing the coprecipitated calcium to about 10 to 30 mg. An important feature of strong nitric acid separations is that lead, barium, radium, and their associated radioactivities also follow strontium, and clean-up steps are necessary before the strontium can be measured for its radioactivity. Much effort has been expended in devising alternative ion exchange procedures allowing selective elution of a pure strontium fraction. Two different approaches are evident. In the first, all cations in the sample are sorbed on a cation exchange column. The calcium and strontium are then separated by fractional elution. Milton and Grummitt (6) in a comparative study of various eluents chose ammonium lactate solution as the most promising for this separation. They developed a procedure for separating strontium from samples containing 2.7 grams of Ca. A sufficiently large column operated at elevated temperatures was used and after most of the calcium had been eluted and discarded, strontium was then eluted nonselectively with HCl and required subsequent radiochemical purification. The author (7) modified this procedure and used it routinely for environmental samples containing up to 3.8 grams of Ca. However, a two-stage separation at elevated temperatures using a large and small column in sequence was necessary in order to obtain a pure strontium fraction from calcium rich samples. The second approach is based on the complexation of calcium with EDTA at a pH at which strontium is mostly cationic. The strontium is then sorbed on a relatively small
x
(2) F. W. Bruenger, D. R. Atherton, and Betsy J. Stover, Healrh Phys., 9, 232 (1963). (3) E. J. Baratta and F. E. Knowles, ANAL.CHEM., 43, 1138 (1971). (4) "Radioassay for Environmental Samples," Environmental Health Series, PHS Publication No. 999-RH-27(1967). ( 5 ) "APHA-Standard Methods for the Examination of Dairy Products", 12th ed., American Public Health Association, 1740 Broadway, New York, N.Y. 10019, in press. ( 6 ) G. M. Milton and W. E. Grummitt, Can. J. Chem., 35, 541 (1957).
(7) L. P. Gregory, Health Pkys., 10, 483 (1964). 2114
cation exchange column while the calcium remains in solution and passes through (8). Strontium is then eluted andpurified, or selectively eluted in a pure form (9). The successful application of this technique depends largely on close analytical control with attendant restrictions, and additional steps such as prior removal of interfering anions by alkaline carbonate fusion, removal of magnesium, and matching the EDTA to the predetermined Ca content. The present paper reports how the difficulties inherent in multiple separations using fuming nitric acid, and in the alternative ion exchange methods, have been overcome by a simple and reliable two-stage procedure: In the first stage, outlined above, the insoluble alkaline earth carbonate or phosphate precipitate is separated from the sample and dried. A single treatment using laboratory reagent 70% HNOI reduces the calcium content to about the same amount as the added strontium carrier. In the second stage, the resulting dilute solution of calcium, strontium, and impurities is sorbed on a small column of cation exchange resin. Fractional elution of the strontium using ammonium lactate solution at room temperature gives radiochemically pure strontium in high yield. A logical development of this method is the procedure for determining the three long-lived bone-seeking radionuclides, *IOPb(tliz 22 yr), ( t i l z 28 yr), and 226Ra (f1i2 1600 yr) in the one sample of bone ash. Advantage is taken of the insolubility of lead, radium, and barium (used to carry the radium) in 70z HN03. The nitric acid separation thus becomes a preparatory step for the sequential separation of these radionuclides from the same column by fractional elution using ammonium acetate, ammonium lactate, and alkaline EDTA eluents, respectively. EXPERIMENTAL
Apparatus. Prepare an ion exchange column by adding 25 ml (wet volume) of 100- to 200-mesh Dowex 50-X8 (NH4+)resin to a borosilicate glass column (14 mm i.d. X 170 mm) connected to a suitable reservoir for feed solutions. Support the resin bed on glass wool to keep the column outlet volume small and thus reduce the intermixing of eluate. Obtain optimum separations during fractional elution by passing about 2 ml of eluent through the resin bed two or three times before filling the space above the bed with the eluent. Reagents. 1.OM Ammonium acetate, pH 6.2. Dissolve 57 ml of glacial acetic acid in about 0.5 1. of water. Add NHIOH solution to pH 6.2 and dilute to 1.0 1. 1.5M Ammonium lactate, pH 7.0. Dissolve 128 ml of lactic acid (88 % w/w) in about 0.7 1. of water. Add NHIOH solution to pH 7.0 and dilute to 1.0 1. (It is preferable to hydrolize lactides by maintaining the diluted acid at 80 O C for 24 hours before neutralization. Concentration and pH constancy are thereby maintained.) 0.25M EDTA, pH 10. Dissolve 93 grams of EDTA disodium salt in about 0.8 1. of water while stirring and adding NaOH to pH 10. Dilute to 1.0 1. Procedure 1. RADIOSTRONTIUM IN VARIOUSMEDIA. Detailed preparation of environmental samples for analysis is not discussed here. Water samples are used as received (rainwater samples do not require nitric acid precipitation, ion exchange separation alone being sufficient). Soil samples are ashed and extracted with HC1. Fe, Mg, Si, and A1 are then separated. The subsequent procedure, on the alkaline earth fraction only, is described here. Food and bone samples are (8) P. S. Davis, Nature, 183, 674 (1959). (9) C. R. Porter, B. Kahn, M. W. Carter, G. L. Rehnberg, and E. W. Pepper, Environ. Sei. Techno[., 1, 745 (1967).
ANALYTICAL CHEMISTRY, VOL. 44, NO. 12, OCTOBER 1972
ashed and aliquots of ash containing about 3 grams of Ca are dissolved in HCl. Strontium carrier (20 mg Sr*+)is added to all samples. When 89Sr measurement is not required, the amount of strontium is increased to 40 mg in food and bone ash samples. Precipitate the insoluble carbonates from water and soil samples with Na2C03. Collect on a glass fiber filter. Dry by washing with methanol and by placing in the oven at 110 OC for a few minutes. Precipitate the insoluble phosphates from food and bone ash by adding H 3 P 0 4and then N H 4 0 H until alkaline. Collect by centrifugation and dry by first dissolving in 20 ml of 70 % H N 0 3and then evaporating to dryness. Add 40 ml of 70% HNOB to the dry carbonate or phosphate. Digest, cool, and filter through a glass fiber filter. Use an additional 40 ml of 70% HNO, to transfer and wash the precipitate. Discard the filtrate. Dissolve the precipitate in 100 ml of water and sorb on the column. Discard the effluent. Elute with 100 ml of 1.5M ammonium lactate, pH 7.0, at a flow rate of about 1 ml per minute. Test the first 15 to 35 ml of eluate for Ca with oxalate solution to determine when all the calcium is off the column. Discard these fractions. Collect the remaining eluate and precipitate SrC03 with ammonium carbonate. Collect and weigh for recovery. Procedure 11. ZlOPb, 90Sr, AND 22eRa IN BONE. Dissolve 10 grams of bone ash in HC1. Add 40 mg Pb 2+, 40 mg Sr 2f, and 5 mg Ba2+carriers. Proceed with the phosphate precipitation, nitric acid separation, and column sorption as in Procedure I. Elute lead with 40 ml of 1 M ammonium acetate, pH 6.2. Precipitate lead with NaHCO,. Collect and weigh as PbC03. Elute and recover strontium as in Procedure I. Elute barium and radium together with 35 ml of 0.25M EDTA, pH 10. Coprecipitate Ba/RaS04 by adding S042- and demasking with acetic acid. All elutions are at a flow rate of about 1 ml per minute. Ion exchange columns are reused repeatedly. Prepare for reuse by first changing to the H+ form with HC1 and then changing to the NH4+ form with ammonium lactate. Wash and backwash with water. Measurement of Radioactivity. Lead-210 is usually determined by measuring the alpha activity of daughter 210P0, separated by spontaneous deposition on silver, after a suitable period for ingrowth. In more active samples, PbC03 may be beta-counted for 210Biafter a shorter period. Strontium-89 is determined by beta-counting SrCO, without delay. Yttrium-90 is then separated after a suitable ingrowth period and beta-counted for 9OSr assessment. Radium-226 is determined by alpha-counting Ba/RaS04 after about four weeks when radium and the three alphaemitting daughters are in equilibrium, thus attaining the counting advantage of 4 alphas per 226Radisintegration.
RESULTS AND DISCUSSION The method outlined in procedure I has been in routine use here for about two years and has been used for the evaluation of international intercomparison samples on several occasions. Strontium recovery is typically 90-95 %. Procedure I1 has been used routinely on human bone samples for about one year. Lead recovery is typically 85%, some lead being lost to the nitric acid filtrate. Strontium recoveries are again over 90 %. Invariably the barium recovery apparently exceeds 100% because of the small amounts of stable barium in environmental samples. Experiments with spiked samples show that radium is quantitatively recovered. Tracer studies show that more than 90% of any 210Poin the ash is removed with the nitric acid filtrate, and that the separated lead fraction is uncontaminated with zlOPo. Ingrowth factors based on pure *lOPbare therefore valid. Experiments also confirm that lead does not interfere in the subsequent deposition of zlOPo on silver. Experiments with heavily spiked samples followed by blank samples confirm that no cross contamination occurs. Procedure I1 is applicable to bone samples because there is no interference by sulfate. For other types of samples, interference may occur through premature precipitation of some BaS04 on addition of barium carrier. The procedure may be extended to such samples by collecting any insoluble BaS04, dissolving in alkaline EDTA, and adding to the EDTA eluate before demasking. Alternatively, alkaline carbonate fusion of the sample ash, followed by recovery and solution of the insoluble carbonates, removes the cause of inter fer ence. Our experience confirms the value of 70 HNO, not only for the separation of strontium from calcium, but also for the separation of lead, barium, and radium. Its use thus allows the subsequent sequential separation of three major long-lived bone seeking radionuclides by simple fractional elution. LLOYDP. GREGORY National Radiation Laboratory Department of Health P.O. BOX25-099 Victoria Street Christchurch, New Zealand RECEIVED for review December 3, 1971. Accepted May 22, 1972. Published with the Authority of The Director-General of Health.
Ultratrace Level Detection of Mercury by an X-Ray Excited Optical Fluorescence Technique SIR: The extensive documentation on the toxic effects of trace levels of Hg present in environmental samples has stimulated interest in the development of analytical techniques for the quantitative determination of Hg at part per billion (1 in 109) levels (I). We report here a promising new technique capable of achieving this objective. (1) R. A. Wallace, W. Fulkerson, W. D. Shults, and W. S. Lyon, “Mercury in the Environment-The Human Element,” ORNL NSF-EP-1, Oak Ridge National Laboratory, Oak Ridge, Tenn.,
1971.
Our studies on X-ray excited optical fluorescence of gases (2) indicated that the optical spectrum of Hg can be effectively excited on X-ray irradiation of Hg vapors in argoc. Further investigations indicated that the intensity of Hg 2537 A was greatly enhanced in a Ar-1 NP (molar %) gas mixture. In addition, intense emission from several molecular Nf band systems was also observed. The most prominent of these (2) A. P. D’Silva, E. L. DeKalb, and V. A. Fassel, Pacific Conference on Chemistry and Spectroscopy, Anaheim, Calif., October 1969.
ANALYTICAL CHEMISTRY, VOL. 44, NO. 12, OCTOBER 1972
2115
