tion with a calibration curve to yield the concentration of the sample. A typical calibration curve for uranyl nitrate solutions in 0.2N nitric acid is given in Figure 5 . The standard absorber was a n ordinary sample container filled with lead acetate solution and sealed with plastic cement to prevent evaporation. The concentration of the lead acetate solution was adjusted to be equivalent in absorptance to about 82.5 grams of uranium per liter. The uranyl nitrate solutions used in these calibrations were obtained by dilution from a stock solution analyzed by standard chemical methods. PRECISION
The ultimate precision of the instrument was evaluated using only one sample vial. An average deviation of 0.1% (apparent standard deviation, 0.05%) at 60 grams of uranium per liter was obtained on the 6- to 8.5-mv. recorder range. This precision was essentially the same for concentrations up to about 200 grams per liter. The error was doubled a t about 30 grams per liter, and doubled again a t about 15 grams per liter. The effect of variations in the diameter of the sample vials on the precision of analyses was also evaluated. One hundred selected vials were filled with a solution corresponding to 60 grams of uranium per liter. An average deviation of 0.6% was found in analyzing these samples with the instrument. The vials had been selected from the common supply by means of a pair of ring gages differing in diameter by 0.003 inch. Approximately 15y0 of the regular stock of vials met this specification. Sorting the vials in this way improved precision by a factor of 3 over that obtained with unsorted vials. ACKNOWLEDGMENT
The information contained in this
Figure 5.
Uranium nitrate calibration curve
paper was developed during the course of work under contract AT(07-2)-1 with the U. S. Atomic Energy Commission, whose permission to publish is gratefully acknowledged. Appreciation is also expressed to B. B. Murray for taking preliminary data which led to this development and to J. M. -1IcKibben for his cooperation in the field evaluation of this instrument. LITERATURE CITED
(1) Bartlett, T. W.,ASAL. CHEX. 23, 705-7 (1951). (2) Hiskey, C. F., Ibid., 21, 1440-6 (1949). (3) Lambert, M. C., U. S. Atomic Energy Comm. Rept. HW-24717 (June 20, 1952).
(4) Leboeuf, h.1. B., Miller, D. G., Connally, R. E., ll’ucleonics 12, No. 8, 18-21 (1964). ( 5 ) Liebhafsky, H. 4., AXAL.CHEW26, 26-31 (1954). (6) Miller, D. G., U. S. Atomic Energy Comm. Rept. HW-39971 (Nov. 17, 1955). (7) Miller, D. G., Connally, R. E., Zbid,, HW-36788 (June 1, 1955) (claw-
fied). Peed, IT. F., Dunn, H. W.,Ibzd., ORNL-1265 (April 8, 1952). Reilley, C. N., Crawford, C. RI., ANAL.CHEM.27, 716-25 (1955). Vollmar, R. C., Petterson, E . C., Petruzzelli, P. A.. ANAL.CHEX.21,
1491-4 (1949). ’ (11) Wright, W. B., Barringer, R. E.,
U. S. Atomic Energy Comm. Rept. Y-1095 (Aug. 26, 1955). RECEIVEDfor review January 17, 1957. Accepted ilugust 6, 1967.
Radioassay of Low Specific Activity Tritiated Water by Improved Liquid Scintillation Techniques C. A. ZIEGLER, D. J. CHLECK, and J. BRINKERHOFF Tracerlab, Inc., Waltham, Mass.
b The performance of liquid scintillators in the radioassay of tritiated water of low specific activity is improved by making the water samplescintillating solution ratio optimum, and deoxygenating the scintillating soh1774
ANALYTICAL CHEMISTRY
tion by an ultrasonic irradiation technique.
T
liquid scintillation technique has achieved considerable popularity as a radioassay method for tritiated HE
water because it combines ease of sample preparation with adequate counting efficiency. However, for samples of low specific activity-i.e., producing a count approaching that of backgroundcounting times become long and the
advantages of the technique become marginal. This paper describes a method of scintillator-sample preparation that increases scintillator performance, resulting in a substantial decrease in counting time for samples of low specific activity. This is accomplished in two ways: by making the water sample-scintillating solution ratio optimum, and deoxygenating the scintillating solution. I n general, the techniques described can be applied in the radioassay of water-soluble radioisotopes other than tritium. MAKING WATER SAMPLE-SCINTILLATOR RATIO OPTIMUM
Because the best scintillating solutions employ toluene, or xylene, as a solvent, it is necessary to use a mutual solvent such as ethyl alcohol to incorporate tritiated water into the solution. [Solutions employing dioxane and naphthalene as well as ethyl alcohol and toluene have been reported ( 5 ) . The general method outlined here for determining optimum ratios of the various solution components can also be applied to such solutions.] The introduction of a tritiated water sample and the mutual solvent ethyl alcohol into a toluene scintillating solution causes a reduction in pulse height and, therefore, in counting efficiency. This “quenching” is primarily due to the
poor energy-transfer properties of water and ethyl alcohol. It is important, therefore, to use the minimum amount of ethyl alcohol required to make a given amount of tritiated water soluble in the toluene scintillating solution. Figure 1 gives an experimentally determined curve of the ethyl alcohol-water ratio necessary for 0.1 to 10.0 ml. of water in total scintillating solution (total scintillating solution refers to 100 ml. of a ternary solution of water-ethyl alcohol-toluene). The ethyl alcoholwater ratio is the minimum recommended value required to maintain solubility a t -25’ C. (in practice most solutions are measured a t low temperatures to minimize phototube noise). Data were obtained using toluene conforming to ACS specifications and absolute ethyl alcohol. -4s a primary scintillating solute 2,5-diphenyloxazole (PPO) a t a concentration of 4.0 grams per liter of toluene mas uqed with 0.1 gram per liter of toluene of 1,4-di-[2-(5phenyloxazole) ]-benzene (POPOP) as a wave length shifter (3, 4). These concentrations in toluene 1%ere used in all the scintillating solutions described. A number of samples using various volumes of tritiated water and the appropriate amounts of ethyl alcohol in a 2,5-diphenyloxasole-1,4-di- [2-(5phmyloxazole) 1-benzene-toluene scintillating solution were prepared and counted by using a Tracerlab liquid scintillation counter, CE-1 ( I O ) . In Figure 2 the counting efficiency is plotted as a function of the water content in milliliters per 100 ml. of total scintillating solution. As the volume of the mater is increased, the counting
efficiency decreases. Samples were counted a t efficiencies up to 12%. The highest counting efficiency, however, is not too meaningful, as solutions that provide the highest counting efficiency are far from optimum solutions. An optimum solution may be defined as the solution that will provide a specified statistical accuracy in the shortest counting time. A small volume of water sample provides high cbounting efficiency but a small number of radioactive events; a large volume of water provides low counting efficiency but a large number of events. Thus, the product of water volume, W, and counting efficiency, E , approaches zero as the water volume approaches either very small or very large values. Therefore, if the product, ETV, of water volume and counting efficiency is plotted against the water volume, the peak of the resulting curve R-ill indicate the optimum volume of water. The parameter values for this optimum solution will depend on the type of scintillating solutes and solvents used, the radioisotope counted, and the efficiency of the counting system used in making the measurements. Thus, the optimum ratio nerd be determined only once for a given solute-solvcnt combination, a given radioisotope, and a given counting system. I n a specified counting system the efficiency depends only on the water volume and not on the specific activity of the solution; thus, the position of the peak of the curve of ETV us. W will not be dependent on spccific activity . The lower curve in Figure 3 is obtained by using 2,5-diphenyloxazole1,4-di- [2-(5-phenyloxazole)]-benzene in toluene solution and tritiated IT ater
16 14
12 10
8
6 4
e I
0
I
I
I
I
I
I
PO
30 40 VOL. OF ETHYL ALCOHOL / VOL. WATER 10
VOL.
1 P 3 OF WATER, ML./100 ML. SOLUTION
0
1 9 3 4 WATER CONTENT. W (ML /700ML
)
Toluene contains 2,s-diphenyloxazole, 4 g./liter, and 1,4-di-[2-
Figure 2. Counting efficiency of airsaturated 2,5-diphenyloxazole-1,4di - [ 2 (5 phenyloxazole)]- benzene in toluene-ethyl alcohol scintillating solutions as a function of tritiated water content
Figure 3. Optimum values of water content for degassed and air-saturated 2,5-diphenyloxazole-l,4 - d i - [ 2 (5 phenyloxazole)]-benzene in solutions
(5-phenyloxazole)]-benzene, 0.1 g./hter). All volumes specified a t 22’ C.
used
Water-ethyl alcohol ratios of Figure 1
Water-ethyl alcohol ratios of Figure
Figure 1. Minimum recommended ethyl alcohol-water ratio for solubility in toluene to - 2 5 ” C.
- -
-
-
1 used
VOL. 2 9 , NO. 12, DECEMBER 1957
1775
counted on a Tracerlab CE-1 liquid scintillation counter. This instrument will always count tritium with a specified statistical accuracy in the shortest time, if a sample volume of 2.3 ml. per 100 ml. of this scintillating solution (not degassed) is used. The use of optimum solutions presupposes the existence of adequate sample volumes. If, however, only a small volume of sample is available (less than the optimum value), the largest available sample should be used. The existence of an optimum does not imply that a sample less than optimum should be diluted u p to the optimum amount. DEOXYGENATION OF SCINTILLATING SOLUTION
The removal of dissolved oxygen from scintillating solutions greatly enhances the light output (2, 8). This improvement has not yet been utilized in the liquid scintillation radioassay of tritium, where any enhancement of light output would be of great advantage. The fact that the tritiated water must remain in solution during deoxygenation means that degassing must occur with negligible solution loss. Furthermore, to be suitable for routine use the degassing method must be simple and convenient. Standard degassing techniques such as vacuum distillation, or bubbling 1% ith a n inert gas, are unsatisfactory, as both methods result in considerable loss of tritium through evaporation and solution carryover. An ultrasonic degassing method for liquid scintillators (1) has been satisfactorily applied to the radioassay of tritiated water. To determine the degree of improvement in light output resulting from the ultrasonic degassing of water-et hy1 alcohol-toluene scintillating solutions, a number of samples of 2,5-diphenyloxazole-1,4 - di [2 - ( 5 - phenyloxazole)]-benzene in toluene scintillating solution containing various volumes of ordinary distilled water and ethyl alcohol TT-ere made oxygen-free. The light produced in these solutions by an external strontium-90 source was determined by measuring the anode current produced in a 6292 DuRlont multiplier phototube. The light outputs of these solutions Tvere compared under identical experimental conditions with light outputs of similar solutions n hich had not been degassed. The per cent increase in light output of degassed relative to undegassed solutions is given in Figure 4 as a function of the water content. I n preparing all solutions the ethyl alcohol-n ater ratios of Figure 1 were used. I n Figure 4 i t is seen that the improvement in light output decreases with increasing n-ater content. This qualitatively corresponds to the fact
-
1776
ANALYTICAL CHEMISTRY
r
,
w
2w 96
g
e4
+
5 22 M
g eo
I
I
I
1
I
P
I
I
I
1 4
3
WATER CONTENT, ML. '1 OOML
Figure 4. Relative increase in light output due to degassing for 2,5-diphenyloxazole1,4-di-[ 2-(5-phenyloxazole)]-benzene in toluene-ethyl alcohol-water scintillating solutions
Kater-ethyl alcohol ratios of Figure 1 used
that oxygen is less soluble in water and ethyl alcohol than in toluene. It is possible for ultrasonic irradiation applied to liquids in the presence of air to cause oxidation reactions in the liquid (6, 7 ) . Peroxides such as hydrogen peroxide, which would be likely to form in aqueous and alcoholic solutions (used in scintillation work), are strong light-quenching agents. To determine if quenchers were being generated, scintillating solutions of varying ethyl alcohol-water content lvere subjected to ultrasonic irradiation times much longer than the time required for degassing. These solutions were then allowed to equilibrate in air, so that the dissolved oxygen content was the same as before irradiation. Sleasured values of the light outputs of these solutions before and after treatment 11-ere the same, indicating that no quenching agents had been generated. Degassed solutions of varying tritiated water content nere prepared and counted. The EW zw. TY curve for these degassed solutions appears as the top curve in Figure 3. Because of the improved performance of the deoxygenated solution, the position of the peak has shifted to the right, so that the new optimum u-ater content is larger. I n practice, by deoxygenation, the counting time to obtain a specified statistical accuracy is reduced by a factor v-hich is the ratio of the EST' values a t the optimunis for degassed and not degassed solutions-Le., 16.3/9.9 = 1.65. This factor holds for source count rates large in comparison to background. For source rates not large compared to background (low specific activity samples), the factor is proportional to the square of the ratios (Q)-i.e.$ 2 7 . The solution loss due to the ultrasonic degassing treatment v a s determined by weighing and found to average 0.757,.
This small loss can be neglected in many applications, and for precise ivork the appropriate weight correction can be easily made. CONCLUSION
Performance of the liquid scintillation method for the radioassay of low specific activity tritiated water can be considerably improved by techniques described. Once the preliminary work of ascertaining the optimum solution ratios has been accomplished, the only additional sample preparation time required is that of the degassing procedure [about 6 minutes per sample ( I ) ] . I n addition to the improvement obtained in scintillator performance by making the solution ratios optimum, an almost threefold reduction in counting time for low specific activity samples can be obtained by deoxygenation. LITERATURE CITED
(1) Chleck, D J., Ziegler, C. -\., Rev. Sci. Insir. 28, 466 (1957). (2) Hayes, F. S . , Los Alamoe Rept. LA-1837 (1O54), unpublished. (3) Haypa, F. S., O t t , D. G., Kerr, i7. S . ,.\-ucleonics 14, 42 (1956). (4) Haves, F. S . , Ott. D. G., Kerr, V. X., Rogers, B. S., Ibid., 13, 38 (1955). ( 5 ) Iiinard, F. E., Reo. Sci. I n s t r . 28, 293 (1957). (6) Miyagawa, I., Research Repts. Fac. Engr., L\-c~yoyaCniv. 3, 31 (1950). (7) Oyama, H., IGsumoto, G., J . Inst. Elec. Enqrs., J n p a n 62, 549 (1012). (8) Pringle, R. W., Funt, B. L., Black, L. D.: Sobering, S., Phys. Ret. 92, 1582 (1953). (9) Thomas, A., Kucleonics 6 , 50 (1950).
(10) Tracerlab, Inc., Tracerlab Catalogue D, p. 76 (1956).
RECEIVEDfor review hIay 4, 1957. Accepted July 29, 1957.
