Radiochemical method for determination of arsenic, bromine, mercury

May 1, 2002 - Rapid, quantitative separation for the determination of selenium using neutron activation analysis. Harry L. Rook. Analytical Chemistry ...
0 downloads 0 Views 467KB Size
aqueous ammonia and was rapidly extracted into propylene of the extractant can be decreased by the addition of certain carbonate. However the addition of EDTA prior to extracelectrolytes. Although three extractions were made in these studies, two tion destroyed the complex. would probably have done just as well, particularly for the The analyses described in this paper were performed to demonstrate that the three iron(I1) complexes could be readily more dilute iron solutions. In all experiments the extracts extracted into propylene carbonate from relatively complex were made to volume with 95 % ethanol. This was probably matrices and be determined spectrophotometrically. unnecessary as they were never cloudy. The A,, and E values shown in Table I are not significantly Propylene carbonate should completely replace nitrodifferent from those obtained in other solvents (5, 8). The benzene as an extractant for the three complexes, particularly shapes of the absorption spectra were also very similar t o the Fe(TPTZ)2+zcomplex. Its density should make it the those obtained in other solvents. Ringbom plots showed choice over the lighter-than-water alcohols for the extraction of the F e ( b a t h o ~ h e n ) ~ +complex. ~ A minor advantage that the three complexes obeyed Beer's law over the range arises because of the low vapor pressure of propylene carstudied. Under conditions of analysis similar to those embonate at room temperature: there is practically no presployed in the recommended procedure, the relative analysis sure build-up when it is shaken with an aqueous solution in a error over the optimum range is 0.6% with the reagents studied when employing spectrophotometers which have a separatory funnel. As is true when using other solvents which extract the three complexes, reagent solutions can be rendered 0.2% photometric error (Table I). iron-free by prior extraction, The p H range over which the complexes studied can be extracted is fairly wide. The range for the F e ( ~ h e n ) ~ + ~ The only disadvantage in using propylene carbonate as an extractant is that it is somewhat soluble in water; however, complex is at least 2 to 9, for the Fe(TPTZ)*'* complex 2.5 to 8, and for the Fe(bathophen)3c2complex 2.5 to 9. These aqueous solutions can be rapidly and predictably saturated with the solvent. ranges are as wide or slightly wider than those reported for the extraction of the complexes into other solvents (5, 8). Propylene carbonate solutions of the three complexes ACKNOWLEDGMENT showed no signs of deterioration over the period of the inThe authors thank Frederick Lindstrom of Clemson vestigation. Also there were n o indications that any of the University for obtaining the absorption spectra and J. C. complexes were precipitating from the solvent. Loftin and Dan W. Olds of Wofford College for technical The solubility of propylene carbonate at 24" C expressed assistance. A donation of propylene carbonate from the as milliliters per 100 ml of liquid was 21.2 in water and 10.4 in Jefferson Chemical Company, Houston, Texas is gratefully a 50:50 mixture of water and saturated sodium chloride acknowledged. solution. The solubility in saturated sodium chloride solution could not be determined accurately by the cloud-point RECEIVEDfor review April 17, 1967. Accepted July 20, method because salt crystals began to precipitate after 5.5 ml of 1967. propylene carbonate were added. Apparently the solubility

Radiochemical Method for Determination of Arsenic, Bromine, Mercury, Antimony, and Selenium in Neutron-Irradiated Biological Material Knut Samsahl' A B Atomenergi, Stockholm, Sweden NEUTRON ACTIVATION ANALYSIS of trace elements forming volatile compounds may in many kinds of materials be simplified by introducing a distillation step. The distillate may then be analyzed by gamma-spectrometry, either directly or after further chemical separations. Detailed methods for the distillation and quantitative determination of groups of radionuclides in different kinds of neutron-irradiated industrial product have been developed by Gebauhr ( 1 , 2 ) and Ross (3). In the present paper a radiochemical method for the determination of As, Br, Hg, Sb, and Se in biological material is described. It is based on a procedure developed for the simulPresent address, Gesellschaft fur Strahlenforschung ni. b. H ., 8042 Neuherberg, Ingolstadter Landstr. 1, Munich, West Germany. (1) W. Gebauhr, Kerntechnik, 4(8), 323 (1962). (2) W. Gebauhr, Radiochirn. Actn, 4 (4), 191 (1965). (3) W. J. Ross, ANAL.CHEM., 36,1114 (1964).

1480

ANALYTICAL CHEMISTRY

taneous distillation as oxides or bromides of 12 trace elements from sulfuric acid solution ( 4 ) . EXPERIMENTAL

Apparatus. The distillation apparatus, shown in Figure 1, is made of borosilicate glass, and ungreased B 14, B 10, or spherical joints are usedas connections. The neck of the distillation flask, A , is surrounded by three turns of a coil with a n inner diameter of 5 mm. This coil, which reduces spray during the distillation, forms the direct connection to the receiver, B. Reagents for flask A are added to a small funnel, H , and then sucked in through a capillary tube which continues along the walls and ends near the bottom of the flask. The distillation flask is surrounded by a borosilicate glass tube in order to ensure sufficient isolation during the distillation. The flask is heated with hot air from a Bunsen burner. (4) K. Samsahl, Aktiebolaget Atornenergi, Stockholm, Sweden, Rept. AE-82 (1962).

H

t l Figure 2. Scheme of anion-exchange separation system Piston barrel, 28 X 150 mm B, C. Piston barrels, 20 X 150 mm D. 5 X 50 mm, Dowex 2 (HSOa-, 200 - 400 mesh) E. 7 X 50 mm, Dowex 2 (Cl-, 200 - 400 mesh) F. 18 X 50 mm (Br-, CI-, 200 - 400 mesh) G . Mixing coils, 5 turns, 15-mm outer diameter H. Piston with rubber stopper 1. Perspex plate, 15 mm thick A.

Figure 1.

Distillation apparatus

Distillation flask, 15-ml volume, 175 mm long, B 14 joint Receiver flask, 30-ml volume, 120 mm long, B 14 joints C. Reflux condenser, 150 mm long, B 10 and B 14 joints D, E. Traps for 2 and 5 ml, respectively, B 10 joints F, H . Funnels, 10 X 100 mm G, I . Stopcocks J . Borosilicate tube, 45 X 380 mm K. Arrow points to water suction pump L . Bunsen burner M. Compressed air A.

B.

The receiver flask, B, is connected in its turn t o a reflux condenser C, two U-tubes, D and E, and a water-suction pump, K . Reagents are sucked into the flask via funnel F. Procedure. The following method is suitable for soft animal tissue samples and may also be used without modification for hard animal tissue containing maximally 6 to 7 mg of Ca. However, it has not been sufficiently tested with plant samples rich in silica. PREPARATION AND IRRADIATION. Maximally 200 mg of a dried, soft animal tissue sample is sealed in a small quartz tube and irradiated together with standards of As, Br, Hg, Sb, and Se for 1 t o 2 days with a thermal neutron flux of 2 X 1013 n/sq cm sec. The high radioactivity of the sample is then allowed to decay for 2 t o 3 days before starting the chemical separation. Further details about the preparation of samples and standards as well as the irradiation are given elsewhere (5). DECOMPOSITION AND DISTILLATION. To flask B are added 1 ml of 1 t o 1 H2SOaand 1 ml of 30z H202. To trap D is added 2 ml of 1 t o 10 H2S04and to E, 5 ml of 6N NaOH. A faint stream of air is now maintained through B, C, D , and E . The quartz tube with the sample is rinsed with acid and water and placed in liquid nitrogen for a short time. The ampoule is then immediately broken. The sample is transferred t o the distillation flask, A , care being taken that glass pieces from the broken tube d o not accompany the sample. Fifty microliters of a 48% HBr carrier solution containing 50 pg each of As+5, Hgf2, Sbt5, and Se+4 is then added and the flask is closed. If nondistillable elements are t o be determined simultaneously, further carrier additions should be made a t this juncture (6, 7). ( 5 ) K. Samsahl, D. Brune, and Wester, P. O., Intern. J . Appl. Rudiurion Isotopes, 16,273 (1965). (6) K. Samsahl, Nukleonik, 8, 252 (1966).

(7) K. Samsahl, Aktiebolaget Atomenergi, Stockholm, Sweden, Rept. AE-247 (1966).

Through funnel H is now added, with suction, 2 ml of 30 to 33% fuming H2SOa. Stopcock I is then closed and G opened, thus maintaining a slow stream of air through the receiver system. The mixture is very carefully heated with a small flame placed a few centimeters below the lower end of the glass tube surrounding the distillation flask. This critical initial heating stage must be watched and regulated t o obtain a n even release of gases without excessive spattering of material into the flask coil. After about 10 minutes the excess of SOn and SOs has been transferred to the receiver. During the next 10 minutes, heating of the contents of flask A is successively increased t o the boiling point. The carbonization is then finished by a short vigorous boiling of the mixture in order to bring down any material spattered around in the flask. The heating is then interrupted and the contents of the flask are cooled to room temperature by means of a blast of compressed air ( I open, G closed). One milliliter of 30% H,Oy is added t o the flask through tube H. When the reaction has ceased (G open, J closed), the mixture is heated to incipient fumes of SOs. Without interrupting the heating, the straw-yellow solution is made completely clear by slowly adding in small drops through tube H about 0.75 ml of 30 H,Oi. Finally, the solution is cooled to room temperature with compressed air. To the wet-ashed sample solution is added 0.5 ml of 48% HBr. To receiver flask B is added 1 ml of 30z H 2 0 r . The additions are followed by strong heating of the solution until boiling, concentrated H2S04 just begins to distill (I closed, G open). After cooling and addition of a new portion of 0.5 ml of 48% HBr to the flask, the distillation is repeated in the same way for a second and a third time. After the third addition of HBr the solution is heated until 0.5 ml of HySOais left in the flask. The distillation procedure is now finished. The 0.5 ml of H S 0 4 solution remaining in distillation flask A contains the greater part of the trace elements present in biological material. Automated ion exchange and extraction chromatographic methods for separating the trace elements into 12 groups suitable for gamma-spectrometric measurements have been described (6, 7). The subdivision of the distilled trace activities as As, Br, Hg, Sb, and Se into four different groups is described below. VOL. 39, NO. 12, OCTOBER 1967

1481

Element As Br Hg Sb Se

Table I. Recovery and Reproducibility of the Method Mean value Standard deviation Standard error Isotope measured of yield, of single value of mean value 76As 90 6 2 82Br 92 7 2 96 1 203Hg 3 lZ4Sb 91 2 1 85 5 ??3e 2

SEPARATION OF BROMINE.While maintaining a faint stream of air through B , C, D , and E the contents of flask B are heated to the boiling point, and heating is continued until the solutions in B and D are completely colorless. This takes 5 to 10 minutes. At this point 95 to 99% of the 8zBractivity has been expelled and quantitatively absorbed in trap E. SEPARATIONOF ANTIMONY,ARSENIC, MERCURY,AND SELENIUM.The trace activities of antimony, arsenic, mercury, and selenium remaining in flask B after the volatilization of bromine are subdivided into three different groups by selective sorption steps in a series of three small anionexchange columns. The separation is performed simultaneously and automatically with a proportioning pump apparatus. The system is schematically shown in Figure 2. The working principle as well as the constructional details of this kind of machine has been described (6). The anion-exchange column, D , is prepared with a few milliliters of 3.6N HYSOlr column E with 4.5N HCI, and column F with a mixture containing equal volumes of 4.5N HC1 and 48 HBr. The piston barrel, A , is filled with 48 % HBr and B with 9N HCI. Barrel C , finally, contains both the sample solution and a washing solution. The sample solution is prepared by adding to the mixture of the acid distillate in receiver B and the contents of trap D 0.25 ml of 8N HC1, followed by dilution to 20 ml with H 2 0 . This solution, about 3.5N in H 2 S 0 4and 0.1N in HCI, is transferred t o barrel C above its upper piston, H , while the washing sohtion consisting of 10 ml of 0.1N HCI occupies the space between the two pistons of the barrel. Plate I is now forced upwards at a predetermined, constant speed corresponding to the delivery of the sample solution t o column D at a rate of approximately 1 ml per minute. The amounts of acid between the columns, injected simultaneously via barrels A and B , make the influent solution about 4.5N in HCI for column E and 2.3N in HCI and 4.5N in HBr for F , respectively. When the sample solution has been forced out of C, a small glass rod in the bottom of the barrel automatically pushes out the rubber stopper in piston H , thus giving free passage for the subsequent washing solution. After about 30 minutes the sample and the washing solution have passed through the system of ion-exchange columns and the machine stops. After the draining of the columns, the resin phases are transferred t o polyethylene tubes, homogenized by means of the corresponding influent solutions, and counted with gamma-ray spectrometry. The bZBr activity in trap E is measured after appropriate dilution with water. Quantitative results are obtained by comparison with the standards

z

(5).

A peristaltic pump may also be used for these separations. RESULTS

Radioactive isotopes of As, Br, Hg, Sb, and Se may be determined in the different groups separated as follows: Group 1. Group 2. Group 3. Group 4. 1482

NaOH absorption: 82Br Dowex 2, sulfate: lg7Hg,za3Hg Dowex 2, chloride: IzzSb, IZ4Sb Dowex 2, bromide, chloride: 7 6 A ~?jSe ,

ANALYTICAL CHEMISTRY

No. of

determinations 6 10 8 7

9

Results of radiochemical recovery and reproducibility studies are given in Table I. The experiments were performed by addition of known amounts of radionuclides to inactive tissue samples, followed by chemical separation and gamma spectrometric comparison with standard samples (8). DISCUSSION

The carriers added to the irradiated material, because of the amount of bromides and chlorides present, may be largely distilled before the destruction of organic material is complete, thus not equilibrating with the total amount of distillable radioactive traces. Accordingly, the radionuclides 7BAs (0.55 mev), z03Hg(0.28 mev), lzzSb(0.56 mev), and 75Se (0.14 mev) were also studied without carrier addition. Suitable compounds of As, Hg, Sb, and Se were irradiated with a thermal neutron flux of about 2 x 1013n/sq cm sec for 5 to 6 days. After a decay period of 2 to 3 days in the case of As, Sb, and Se and about 3 weeks in the case of Hg, gamma peak-height activities corresponding to 50 to 100 counts per minute close to a 3 x 3 inch solid NaI crystal were used for the distillation experiments. These were for each radionuclide carried out, partly after the addition of about 100 mg of dried, inactive kale powder to the distillation flask, and partly without the use of a biological sample. The subsequent gammaspectrometric measurements of the contents of the distillation flask, the emptied flask, and the distillate showed that no activity was adsorbed on the glass walls, and all the elements distilled 95 to 100 quantitatively, independently of carrier addition. This is in agreement with the results obtained by Ross with pure industrial products (3). However, the small amount of carriers used in the present method, as well as in the subgrouping procedures described earlier (6,7), will not cause complications-e. g., due to precipitation or colloid formation in the solutions, Their use will in any case be advantageous, as recovery studies and test experiments can then for the most part be carried out with strongly radioactive tracers, thus largely simplifying counting operations and calculations. The final volatilization of bromine activity from the distillate can be performed almost quantitatively only through the use of a comparatively large amount of bromine carrier throughout the whole procedure. A probable explanation of this is that parts of the distilled activity are adsorbed as elementary bromine to spattered microamounts of organic material in the receiver flask. The present method has been tested with NaCl added in a n amount corresponding to 20 mg of chloride ions. There is still no risk that volatile chlorides of As, Hg, Sb, and Se will accompany bromine to the NaOH absorption trap. The bromides of these elements, are destroyed in receiver B, because of the excess of HpOnpresent. The method may consequently be regarded as completely reliable for biological material containing chloride ions in the range 0 to 20 mg.

z

(8) K. Samsahl, P. 0. Wester, and 0. Landstrom, Stockholm, unpublished data, 1966.

The main disadvantage of the present method, and of distillation procedures in general, is the difficulty of introducing automatized operations. In the serial analysis of strongly irradiated biological samples this is a drawback, owing to the high radiation levels, among other things. Some improvements on the present method along this line might include the fast destruction of organic matter with NalOz fusion and subsequent distillation of bromides from acid solution with a continuous stream of dry HBr gas.

ACKNOWLEDGMENT

The author is greatly indebted to Erik Haeffner, head of the Chemistry Department, for his active interest in this work and to Sigrid Hackbarth and Solveig Hellman for skillful technical

assistance,

RECEIVED for review December 5 , 1966. Accepted May 4, 1967.

Modification of the Fluoride Activity Electrode for Microchemical Analysis Richard A. Durst and John K. Taylor Institute f b r Marerials Research, Dicision of Analytical Chemistry, National Bureau of Standards, Washington, D. C. 20234 THEFLUROIDE ELECTRODE ( I ) consists of a laser-type singlecrystal membrane (europium-doped lanthanum fluoride) which responds t o fluoride activity and is sealed into the end of a rigid polyvinyl chloride tube. Inside this tube and in contact with the inner surface of the membrane is a NaF-KCl solution. A silver/silver chloride wire serves as the inner reference electrode and lead wire contact. In normal operation, using a saturated calomel reference electrode, measurements can be obtained on as little as one or two milliliters of sample solution. Because the bulky electrode tube is the major limitation on sample size by preventing the immersion of the crystal membrane in smaller volumes, the electrode must be modified for use with sample volumes of less than a milliliter. Operation of the electrode in the inverted position makes possible the use of the membrane itself as the sample “container’’ and permits measurements on volumes as small as one drop. However, on inversion of the electrode, the entrapment of an air bubble next t o the inner surface of the membrane may cause erratic operation, or complete failure of the electrode if. the solution-membrane contact is broken. To alleviate this problem, it is possible to completely fill the inner compartment of the electrode with the reference solution t o eliminate the dead air space. However, the inner construction makes complete filling of the electrode difficult, so that an alternate technique is preferred. This consists of filling the electrode with a gel composed of 4z agar, O.1M NaF, and 0.1M KCI. To accomplish this, the electrode compartment is opened (screw top and inner ring seal) and heated by placing it in a beaker of water on a steam bath. The hot agar/salt solution is then injected into the small inner compartment of the electrode by means of a syringe. The electrode is reassembled and allowed t o cool slowly in the normal vertical position. By using this technique, complete filling of the inner compartment is not necessary since any air remaining above the gel will not cause improper behavior on inversion of the electrode because air bubbles cannot rise through the gel t o block the membrane. After the electrode has been modified for use in the inverted position, it is necessary to confine the sample solution to a (1) Orion Model 94-09 Fluoride Activity Electrode, Orion Re-

search Inc., Cambridge, Mass. [M. Scierice, 154, 1553 (1966)l.

s. Frant and J. W. Ross, Jr.,

S.C.E.(FIBER JUNCTION1

SINGLE- CRYSTAL FLUORIDE-SPECIFIC MEMBRANE ,-SAMPLE

SOLi;TlON(5$pl)

TYGON SLEEVE

MODIFIED

FLUORIDE ACTIVITY ELECTRODE

L lOmm--(

Figure 1. Modified fluoride electrode microcell and SCE

fixed area of the membrane and t o prevent the spreading of a solution film, where the fluoride activity may be different from that in the bulk solution, t o other parts of the fluoride sensitive membrane. This is accomplished, as shown in Figure 1, by the use of a Tygon tubing (plasticized PVC, 6-mm i d . ) sleeve which forms a very tight seal with the raised portion of the membrane. To prevent possible solution entrapment and carryover of sample solutions which would reduce the electrode response, a hydrophobic barrier in the form of a film of silicone lubricant is applied between the crystal membrane and the sleeve. To complete the cell, the test solution is pipetted onto the membrane, and a fiber junction saturated calomel electrode is lowered into contact with the solution surface (Figure 1). Although 50 pl were used in evaluating the response of this modified electrode, solution volumes as small as 25 p1 are VOL. 39, NO. 12, OCTOBER 1967

1483