LITERATURE CITED
(1) Bavin, P. M. G., ANAL.CHEM.32, 554 (1960). (2) Bodroux, F., Compf. Rend. 138, 92, 700 (19041. (3) Bo;vea&, M. L., Bull. SOC.Chim. 31, 1322 (1904). (4) Bouveault, M. L., Compt. Rend. 137, 987 (1903). (5) Boyd, R. N., Meadow, hf., ANAL. CHEM.32, 551 (1960). (6) Fales, H. M., J . Am. Chem. SOC.77, 5118 (1955). (7) Gatterman, H., Maffezzoli, I., Be?. 36, 4152 (1903).
(8) Gilman, H., Schulze, F., J . Am. Chem. SOC.47, 2002 (1925). (9) Jones, R. G., Ibid. 69, 2346 (1947). (10)Kharasch, M.. Reinmuth, O., ‘‘Grignard Reactions of Nonmetallic Substances,” p. 75, Prentice-Hall, New York, 1954. (11) Monier-Williams, G. W., J . Chem. Soc. 89,273 (1906). (12) Nichol, J. G., Sandin, R. B., J . Am. Chem. SOC.67, 1307 (1945). (13) Sharefkin, J. G., Saltzman, H., ANAL.CHEM.35, 1428 (1963). (14)Shriner, R. L. Fuson, H. C., Curtin, D. Y., “The Systematic Identification of Organic Compounds,” 4th ed., Chap. 9, Wiley, New York, 1956.
k,
(15) Smith, L. I., Nichols, J., J. Org. Chem. 6, 489 (1941). (16) Tschitschibabin, A. E., Ber. 37, 186 (19041. --,~ \ - -
(17) Ibid., p. 850. (18) Wuyts, H., Ber. 38, 195 (1929). (19) Ibid., 39, 58 (1930). (20) Ibid., 40, 665 (1931). (21) Ibid., 41, 196 (1932). RECEIVEDfor review April 17, 1963. Accepted July 2, 1963. Division of Analytical Chemistry, 144th Meeting, ACS, Loa Angeles, April 3, 1963. From B thesis submitted b Alex Forschirm
in partial fulfillment o f the M.A. degree, Brooklyn College, Brooklyn, N. Y., February 1963.
Determination of Beryllium by the Photoneutron Method GERALD GOLDSTEIN Analytical Chemistry Division, Oak Ridge National Laboratory, Oak Ridge, Tenn.
b Apparatus was designed and constructed for the determination of beryllium by the photoneutron method which can be employed in any analytical laboratory without hazard or difficulty. Because large BloF3 neutron counter tubes (2-inch) are used, the neutron detection efficiency of the instrument is sufficiently great so that a gamma source activity of only -300 mc. of Sb124 is necessary for most samples. With this source activity sample solutions containing as little as 60 pg. of Be per ml., or 1.5 mg. of total beryllium, can be analyzed with a relative standard deviation of =tl Interference caused by elements which absorb neutrons can be eliminated by the cadmium shield technique. Deuterium is an interference if present in a 25-fold excess over beryllium. Results are shown for the determination of beryllium in several types of materials.
yo.
S
the photoneutron method for the determination of beryllium was first introduced (8, 9), it has been extensively utilized to detect and measure beryllium in geological samples (1-5, 17, 18,do), ore processing samples (15), and even in beryllium metal (19). Portable instruments, which have revolutionized beryllium prospecting methods, have been designed (1, 2, 4, 10, 12, 14, 17, 18, 20) and are commercially available. I n most respects the photoneutron reaction provides an ideal analytical method; it is rapid, nondestructive, and virtually free of interferences, and is far more versatile and reliable than the chemical methods used for beryllium analysis. This method is, therefore, potentially of great value as a laboratory technique because it is capable of determining beryllium over a wide coneen? INCE
1620
ANALYTICAL CHEMISTRY
tration range and in a large variety of materials without prior separation of the beryllium. Instruments for applying the photoneutron method in the laboratory have been described by British (f3, 15), Russian (14), French (6), and Italian (7) workers, but these have the disadvantages that large samples are required and a gamma source of high activity (1 to 1.5 curies) is necessary for adequate sensitivity. Consequently, work was undertaken to design and construct an instrument suitable for use in the average analytical laboratory. The primary objective was to devise an instrument which is simple to construct and operate, can be used for both solid and liquid samples, and provides adequate sensitivity withminimum radiation hazard. I n addition, for our purposes, a certain amount of portability was desirable-Le., when the source is removed, the instrument can be transported from place to place. Excellent discussions of the principle of the photoneutron method and the considerations involved in designing an instrument are available (6, 6, 11, 18, f 4 ) arid will not be repeated here. EXPERIMENTAL
Apparatus. A schematic diagram of the apparatus is shown in Figure 1, and details of the sample stage and container are presented in Figure 2. The electrical connections to the BF, tubes are connected by a single concentric cable t o the high-voltage supply and pre-amplifier. Associated electronic equipment consisting of a high-voltage power supply (NJE Corp., Model 5-324), linear amplifier (Atomic Instrument Go., Model 2lS), scaler (Atomic Instrument Co., Model 101B), and timer are placed in a 4foot rack. Four inches of lead shielding was stacked around the apparatus which was placed on a table in a corner of the laboratory, and the table area was roped off.
Reagents. Standard beryllium solution, 10.0 mg. per ml., was prepared by dissolving 5.00 grams of beryllium metal in dilute HCl, then diluting the solution to 500 ml. with 10% HC1. Other solutions were prepared by appropriate dilution of the stock solution. Procedure. INITIAL ADJUSTMENT. With the Sblz4source in place, set the band width a t 0.5 megacycle, high voltage a t 1600 volts, am lifier coarse gain to 16, fine gain to 1, pu se height selector a t maximum, and count 1-minute backgrounds a t decreasing pulse height discriminator settings. Stop when the counting rate becomes greater than 100 c.p.m. Then increase the high voltage by 100 volts and repeat, until 2000 volts is reached. Insert a sample in position (2 mg. of Be per ml. in 5ml. cell) by loading the sample on the sample stage and slipping the stage over the source rod by means of %foot tongs, and obtain counting rates a t the same settings as the backgrounds. From this data, instrument settings can be chosen a t which the net sample counting rate is a t a maximum and is independent, or almost independent, of high voltage and pulse height discriminator settings. If there is any doubt that the initial gain settings are the best, repeat the adjustment procedure a t other settings. This adjustment should be repeated once a month. CALIBRATION.Obtain net counting rates for standard beryllium solutions as follows. I n the 5-ml. cells count standards ranging from 1.0 to 3.0 mg. per ml.; in the 25-ml. cells, 0.3 to 1.0 mg. per ml. standards; and in the 50-ml. cells, 0.1 to 0.5 mg. per ml. standards. Collect a t least 10,000 counts in each case. Calculate an average calibration factor for each cell (c.p.m. per mg. per ml.). SANPLEANALYSIS. Pipet a 5-, 25-, or 50-ml. portion of a sample solution containing between 0.1 and 3 mg. of beryllium per ml. into the ap-
P
4
' /I 7 1
Figure 1 .
Schematic diagram of apparatus
Sbl*'source, 40 mg. of 98% enriched Sbl*a sealed in a quartz tube and packed in a 1 l/* X 0.25-inch screw-cop aluminum can b. Source holder, 6/~e inch 0.d. X 1 1 -inch brass tube fixed to a circular base plate 1 inch diameter X I/( inch thick C. Sample coniainer d. l e a d cylinder, 4 inches in diameter X 12 inches in height avtrall, with a well 2 inches in diameter X 7 inches deep. e. Moderaior, 1 2 X 12-inch polyethylene cylinder with a cenlrol hole 4 inches in diameter, and 5 equally spaced 2-inch holes drilled a t a radius0f;3~/2 inches. f. B10F3 neutron counter tube. 5 tubes, 2 inches In diameter X 12 inches over-all length, 40-cm. pressure. Reuter-Stokes RSN20A g. Desiccated space containing BFs tube electical connections a.
I
propriate cell. Load the sample in the apparatus and count until a t least 10,000 counts are collected. Calculate counts per minute and with the calibration factors determine milligrams of Be per milliliter. RESULTS AND DI:;CUSSION
Evaluation of Apparatus. Although this apparatus, as well as other instruments designed for use in the laboratory, is fundammtally the same as Gaudin and Pannc,ll's original (Q), several improvements have been incorporated. The use of enricht:d SbIZSas the source, as suggested by Milner, Edwards, and Henry (IC), rather than natural antimony haai several advantages. Thermal neutron activation of natural antimony proc uces high activities of SblZ2(Table I:i which takes no part in photoneutron generation. I3esides increasing the radiation hazard, the additional gamma flux can cause instrumental difficulties. Only a small weight of enriched SblZais necessary to produce sufficient Sblz4 activity; 40 mg. of 98% enriched SblZ8irradiated for 14 days in the OaE. Ridge Research Reactor produced about 400 mc. of Sb124. Polyethylene is used as the moderator rather than paraffin be:ause of its better physical properties. Increased sensitivity was achieved by the use of 2-inch BFa tubes instead of the 1-inch tubes normally employed. The larger tubes are about 5 times more efficient than the smaller. I n tests with a calibrated Arn241-& neutron source (4.5-m .e.v. average energy), the efficiency of the apparatus was 5.1%. However, the backgroLlnd is aIso greater, 40 c.p.m. compared to -20 c.p.m. for 1-inch tubes, and the 2-inch tubes are somewhat more sensiti7re to gamma radi-
ation. Because of the effect of gamma radiation on the tubes, a high voltage plateau may not be found when the instrumental operating conditions are evaluated (11, 14). Figure 3 shows typical data which were obtained in the initial adjustment procedure with a beryllium solution in place, Most regulated high voltage supplies are suffi-
Table 1. Calculated Activities" of Sb122 and Sb1*4 in Natural Antimony and 98% Enriched SblS after 14-Day Irradiation
Cooling time, days 0
7 14
Sb1Z4, mc.
Sb122, mc. Natural 98% Sb SbI23
31 - -5 _
290 268
61 11 1.9
4220
747 131
I, Does not include resonance activation.
Table II. Approximate Radiation level at Various Positionswith a 300-400 mc. Source
Position Under table Surface of Pb barrier Top of Pb barrier Top of instrument Top of Bource holder
Figure 2. container
Sample stage and sample
a. Source holder b. Sample stage. Brcss tube, 8/a inch 0.d. X 14 inches in height fixed to a circular 1 inch diameter base plate having a centrol hole 3/8 inches in diameter C. Sample container, fused silica. Bate plate 1 a/d inches in diameter X I/* inch thick having a central hole 1 cm. in diameter. Inner tube 10 mm. i.d. X 1 3 mm. 0.d. X 3-inch height. Outer tube 3-inch height X 16, 25, or 3 5 mm. i.d. for capacities of 5, 25, and 50 ml., respectively d. Sample
mr. /hr. 20 2 . 2 m.e.v. (Coj6, for example), sensitivity could probably be increased by two orders of magnitude. Application to Samples. T h e apparatus described has been used t o determine beryllium in various types of samples as shown in Table VI, compared with results obtained b y colorimetric methods after appropriste separations. I n all cases the sample was first p u t into solution by the most convenient method and a portion of the solution analyzed directly withcut any separations by the recommended procedure. Counting times required were not greater than 10 minutes \rith a 300-me. source. The relative standard deviation of duplicates was i1.170, slightly smaller than the theoretical value based on counting statistics. CONCLUSIONS
a laboratory method, the photoneutron method has the advantages of rapidity, versatility, and high precision compared to most chemical methods for beryllium analysis. Chemical separations are rarely necessary. Some colorimetric and fluorimetric methods, however, are more sensitive than the photoneutron method as described here. I t i5 probably not practical in an ordinary laboratory to increase the seiisitir ity of the photonputron method hv iiying more than 500 me of Sb124 Thr d e c w ~ of the iouice can also be a dranbark i f beryllium determinations are only performed at irregular intervals. Some mention should be made of thr cost of this apparatus. Materials, including elevtronics and enriched SblL3, are about $3000 Irradiation charges are $150 for 14 da>b, and for most purp o b ~ sthis 400-nic. source is useful for about 2 half-lives (120 day>) Sb124 sources are also commercially a\ aiiablc
(9) Gaudin, A. M., Pannell, J. H., ANAL. CHEM.23, 1261 (1951). (10) Hale, F. H., Bishy, H., "Radio-
LITERATURE CITED
. .
(1960):' ' ( 2 ) Bisby, IT., I'.K . Ai'. Energy Authorily Rept. AERE-R-3021 :1Y59). (3) Bowie, 8. H. U., 'Bisby, H., Burke, I
