- Email: [email protected]
Measurement of 99Tc in low-level radioactive waste from reactors using 99mTc as a tracer
Appl. Radiat. Isot. Vol. 38, No. 6, pp. 447450, ht. .I. Radiat. Appl. Insrrum. Parr A Printed in Great Britain. Ail fights reserved
1987 Copyright
6
0883-2889/87 $3.00 + 0.00 1987 Pergamon Journals Ltd
Measurement of 99Tc in Low-Level Radioactive Waste from Reactors using 99mT~as a Tracer J. E. MARTIN* School
of Public
Health,
University
and
of Michigan,
J. M. HYLKO Ann Arbor, MI 48109-2029, U.S.A
(Received 20 October 1986; in revised form20 November 1986) waste procedure has been developed for measuring 99Tc in major low-level radioactive streams from commercial nuclear power facilities which contain numerous potential contaminants. Separation was by alkaline precipitation and anion exchange chromatography; trace amounts of “Co and other radioactive contaminants were removed by solvent separation. Pure separations were obtained with average radiochemical yield of 47.3 5 4.3% determined by a 99mTctracer. The lower limit of detection for
A radiochemical
%Tc was found to be 0.16 pCi/g based on IO-g sample and a IOO-min counting time.
Introduction Technetium-99 is an important constituent of low-level radioactive waste (LLW) and has received particular attention as a source term for shallow land burial sites (U.S. NRC, 1982). Regulations promulgated by Title 10, Part 61 in the Code of Federal Regulations (10 CFR 61) require concentrations of p9Tc to be reported on LLW shipping manifests so site inventories can be established. This requirement is basic to determining the potential impact of any new sites operated by regional LLW compacts in accordance with the Low-Level Waste Policy Act of 1980 (U.S. Congress, 1980). Technetium-99 has a half life of 2.13 x 10’~ (Kocher, 1981); however, Coursey et al., (1984) in recent studies suggest that 2.111 + 0.012 x 10’~ may be a more appropriate value. The nuclide is a pure beta-particle emitter with a maximum beta energy of 292 keV. Its primary mechanism of production is by thermal neutron fission of 235U and 239Pu in nuclear reactors; thus, nuclear power plants represent a significant source of 9qc. Production of 9qc also occurs at a low rate by neutron activation of stable 98M~ which exists in fuel as a stable fission product or outside fuel as a trace constituent in reactor structures (Till et al., 1985). Various analytical procedures have been used to measure ‘%I’c in samples. Chu and Feldstein (1984) list several such methods and their limitations for measuring 99Tc in environmental samples using betaemitting Tc tracers to measure chemical recovery. Walker et al. (1980) report the use of liquid scintillation counting as a quick and inexpensive technique for measuring !@Tc if chemiluminescence is carefully * Author
for correspondence.
eliminated, and Coursey et al. (1984) have used it as a means of standardizing 99Tc solutions. Foti et al. (1972) have used neutron activation analysis to measure 99Tc at the picocurie level. The induced ‘“Tc activity was measured by y-ray spectroscopy immediately after irradiation. Since the half life of the lWTc is only 16 s, computerized data acquisition and analysis are required to process the y-ray spectra. Technetium can also be separated by distillation and solvent extraction of aqueous-organic phases. The principal disadvantage of this method is the introduction of acid or organic mixtures to reduced the pertechnetate (TcO;) ion which causes difficulties in subsequent separation steps using ion-exchange chromatography (Anders, 1960). The feasibility of 99Tc analysis by gravimetric, spectrophotometric or polarographic techniques has not been demonstrated since these methods are not sufficiently sensitive for environmental samples (Till et al., 1985). In general, a reliable procedure for measuring 9?c in the diversity of LLW samples from reactors has not been reported. Since the primary source of 99Tc in LLW in nuclear reactor waste streams, it is important that the method be applicable to quantitating the concentrations of 9qc in the various liquid and solid media such wastes comprise. Because such waste streams contain numerous beta emitters that could interfere with the accuracy of 99Tc measurements, it is important to assume that the measured concentrations represent the 9‘Tc and not a contaminant artifact; thus, pure separations are required. The objectives of this study were to develop such a procedure, to establish a lower limit of detection for it, and to determine the value of using 99Tc as a tracer to measure chemical recovery. Although these studies did not consider 9qc in other non-fuel-cycle LLW, such as that from research and medical institutions, 447
J. E. MARTIN and
448
the same methodology and approach be applicable for these wastes.
are believed to
Method Samples of LLW types representing about 6 months operation were obtained from the Big Rock Point and Palisades Nuclear Plants in Michigan. Detailed summaries of the types of samples analysed have been reported for 14Cfor these plants and others by Martin (1985). Other data on 9’Tc are presented by Hylko (1986) on studies of the effect of wall absorption in polyethylene containers on measurements of 99Tc in liquid waste samples. These studies showed that about 50% of the 99Tc absorbed onto the container, but that this could virtually be eliminated by sample pretreatment with paper pulp and HCI. This sampling feature is incorported into the analytical method now used for 99Tc determinations in LLW. The radiochemical procedure reported by Chu et al. (1984) was modified for separation of 99Tc from mixed activation, corrosion, and fission products. Liquid samples containing paper pulp for pretreatment were filtered through a 0.2 pm filter to remove particulate matter so the filtrate and filterable materal could be analyzed separately. A measured volume of 99mTc extracted from a 99Mo-99mTc generator was added prior to the chemical separation as a tracer to determine chemical recovery. A 2 x 2-in. NaI(TI) crystal was used for counting the 99mTc tracer; it was not necessary to standardize the tracer since a ratio method was used on identical volumes added to each sample and counting before and after separation. The activity of the tracer spikes was low enough to minimize the effect of dead-time losses. Solid samples were dissolved under reflux conditions in 7.5 M HNO,. If solid material remained, the sample was filtered and the solid material was refluxed a second time and analyzed separately. Five milliliters of Sr” and 5 mL of CO’+ were added as carriers; each contained a nitrate concentration of 20 mg/mL in solution. The samples were then treated with NaOH and Na,CO, to precipitate the actinides, cesium, strontium, and the transition metals leaving 99Tc in solution. The filtrate was acidified by adding 7.5 M HNO, to the solution until the orange-colored methyl red endpoint was achieved, then 12 mL more of the acid was added which changed the solution color to a clear pink. The solution was heated gently to remove the CO, produced by the acidification of the carbonate, and the solution was diluted with deionized water to 870 mL yielding a final solution containing about 0.1 M nitric acid. The 99Tc solution was concentrated on a Dowex 1 x 4 anion exchange resin further separating it from other trace nuclides present in the weakly acidic solution. The resin was washed with 500 mL of 0.1 M HNO, and the concentrated 99Tc was eluted from the resin at full flow with 100 mL of 5 M HNO, at a
J. M.
HYLKO
recovery of about 90%. Evaporation temperatures were held between 80 and 90 C for the column elutions to reduce sample loss. which was less than 10% under such conditions. The eluate pH was increased to between 10 and 11 using concentrated NH,OH, followed by addition of 10mL of 0.5 M tetraphenylarsonium chloride. The mixture was transferred to a separatory funnel containing 100 mL of chloroform and was shaken for 2 min. The chloroform fraction was added to a second separatory funnel containing 50 mL of 0. I M NH,OH and was shaken for 2min. The NH,OH fraction was saved and the alkaline wash was repeated. The solution was evaporated to less than 10mL and transferred to a ringed stainless-steel planchct and evaporated to dryness. A low-background gas-flow proportional counter was used to measure the separated samples. The counter was calibrated for 99Tc radioassay using a solution of “Tc standard (Amersham Corp., Arlington Heights, Ill.) evaporated on ringed stainless steel planchets. An identical volume of unprocessed 99mTc spike (100% recovery) was counted followed by a processed samples with a Nal(T1) crystal and a single channel analyzer for chemical recovery. Total counting time was varied to reduce counting error to less than 5%. The net counting rate of the 99mT~in each sample was corrected for decay to the time the unprocessed spike was counted in order to determine chemical recovery. The samples were stored for a minimum of 60 h for 99mTc to decay before counting for 99Tc. The 99Tc activity per gram of sample, A (99T~), was calculated by: A (“Tc) =
C net EWY
2.22 x IO’* d/mCi
where C,,, is the net beta count rate (cpm) of the sample with background subtracted (based on a counting time of lOOmin), E is the 99Tc counting efficiency, W is sample weight, and Y is the chemical yield based on the decay-corrected count rate of the y9Tc in the processed sample divided by the count rate of the 99mTc spike (100% recovery). A lower limit of detectability based on the method of Currie (1968) was determined.
Results and Discussion The experimental procedure used was orders of magnitude more sensitive than the level of 0.03 Ci/m’ required by 10 CFR 61 for LLW waste characterization (U.S. NRC, 1982). Considerable work was done to establish the procedure and verify its yield, efficiency, and reproducibility. Table 1 contains the results of analyses made with 99Tc and 99mT~tracers to establish recovery, which was found to be 47.3 + 4.3% for liquid samples through the precipitation, anion-exchange, and solvent separation procedure; solid sample recoveries were slightly better. Memory checks were run following each processed standard
%Tc in reactor Table
that complete equilibrium of the 99mTctracer and 9qc in the samples was accomplished independent of the sample matrix. This was confirmed by analyzing two sets of identical samples.
1. Chemical recovery for Drocessed samoles Sample
Recovery (%)
Reactor water (treated) Reactor water (untreated) Concentrates Resins Crud Waste filter Smears
45.01 47.8’ 46.5’ 48.57 48.5t 56.0 45.0
449
wastes
Sample con taminan ts
l Average of recoveries for filtrate and the filter. able fractions for two samples. t Average of two samples.
and in the course of the experiments to establish that residual deposition in the apparatus was indeed insignificant. Background counts with blanks were about 1.4-1.6 cpm, which with a lOO-min counting time yielded a minimum detectable activity (Currie, 1968) of about 0.16 pCi/g for a 10 g sample at 47.3 k 4.3% recovery. Solid samples The method was applied to solid samples by refluxing the sample in 7.5 M nitric acid. This refluxing procedure proved to be efficient for uniform dissolution of the samples. Solid material that did not dissolve during initial pretreatment was analyzed a second time; however, sample activity was always equivalent to background levels. When excess refluxing was necessary for dissolving such samples, some chemical recoveries dropped below SO%, demonstrating the value of the 99mT~tracer for non-liquid samples. Identical samples analyzed a second time from the same or similar containers were, with 95% confidence, statistically equal to the initial 99Tc activity that was measured. This strongly suggests
Because LLW samples contain significant radioactivity from numerous nuclides, the presence of contaminants in the %Tc samples is always possible even after chemical separation. Cobalt-60 is the most common nuclide found after radiochemical separations, probably because it represents the highest measured concentration in most LLW samples and because of its chemical properties. Decontamination of 99Tc from trace amounts of ‘j”Co and other contaminants was achieved by the solvent separation step. In order to establish the 99Tc was in fact separated from higher energy beta-emitters, a HarleyHallden plot (Volchok and de Planque, 1982) was determined for each sample. The slopes of the plots were determined to be statistically equal to unity at the 95% confidence level (Hylko, 1986). Liquid scintillation counting was also used to determine the presence of other beta emitting contaminants. Figure 1 shows a plot of the beta spectra of a 99Tc standard and two LLW samples (without quench correction) determined with a Beckman LS 6800 Liquid Scintillation Analyzer (Beckman Inc., Irvine, Calif.) which has a logarithmic amplifier. The spectra of the two LLW samples show comparable energy distributions with that of the 9qc standard, providing additional confirmation of the sample purity. Since 6oCo contamination (maximum beta energy of 0.317 MeV) is a likely contaminant that could by masked by 99Tc in liquid scintillation counting, the samples were also gamma counted using an
l.O-
‘.._4-
.... .il
-0
I&
260
300 CHANNEL
Fig. 1. Liquid
scintillation
460
So0
“To
Reactor
Concentrates (H+213)
SbO
NUMBER
analyzer (Beckman Model LS 6800) beta spectra of two ‘Vc separations reactor waste samples compared to a WTc standard.
of
J. E. MARTINand J. M. HYLKO
450
intrinsic germanium detector on the presumption that beta-emitting contaminants in LLW would also emit y rays. Such contaminants would distort the proportional counter results and liquid scintillation analyses due to the spectral distribution of beta energies. All beta and gamma spectra of processed LLW samples indicated that the samples were free of radionuclide contamination and were equivalent to processed 99Tc standards.
Summary and Conclusions This study was undertaken to develop a methodology for measuring 99Tc in various liquid, solid, and semi-solid low level radioactive waste samples. A modification of Chu’s method (Chu and Feldstein, 1984) using a combination of alkaline carbonate precipitation, anion-exchange chromatography, and solvent separation provided excellent isolation of 9’?c from mixed activation, corrosion, and fission products for the usual forms of reactor LLW that require classification. Good separation of qc was obtained with yields of about 47.3 f 4.3% and the method eliminated any significant interference by the myriad of other constituents in the LLW samples. Technetium-99m, the pure gamma-emitting isomer of ‘?c, was found to be an excellent tracer for determining chemical recovery because its short half life (6.007 h) allowed complete decay prior to counting for 9qc. Low background beta counting of processed samples allowed reliable detection of ?c concentrations in LLW samples of 0.16 pCi/g based as a 10 g sample and 47.3% chemical recovery.
References Anders E. (1960) Radiochemistry qf Technetium, NASNS3021 National Academy of Sciences-National Research Council, Washington D.C. Chu N. Y. and Feldstein J. (1984) Radiochemical determination of technetium-99. Talanta 31, 809. Coursey B. M., Gibson J. A. B., Heitzmann M. W. and Leak J. C. (1984) Standardization of technetium-99 by liquid scintillation counting. ht. J. Appl. Radial. Isot. 35, 1103. Currie L. A. (1968) Limits for qualitative and quantitative determination. Anal. Chem. 40, 586. Foti S., Delucchi E. and Akamian V. (1972) Determination of picocurie amounts of WTc by neutron activation. Anal. Chim. Acta 60, 261. Hylko J. M. (1986) The Analysis of %Tc and “‘CS in Low -Level Waste Samples from- Commercial Nuclear Power Facilities. Master’s Thesis (Unuublished). School of Public Health, University of Michigan, Ann Arbor, MI 48109, U.S.A. Kocher D. C. (1981) Radioactive Decay Data Tables. DOG/TIC 11026, 108. Martin J. E. (1985) Carbon-14 in low-level radioactive waste from two nuclear power plants. Health Phys. 50, 57. U.S. Nuclear Regulatory Commission (1982) Licensing requirements for land disposal of radioactive waste. Federal Re.&ter 47, 51446 (27 December 1982). Till J. E., Sh& R. W. and Hoffman F. 0. (1985) Uranium Fuel Cvcle-A Review of Data for Technetium. NUREGI CR-3738. U.S. Nuclear- Regulatory Commission, Washington, D.C. U.S. Congress (1980) Low-Level Waste Policy Act qf 1980, Public Law 96-573, 94 Stat. 3347. U.S. Government Printing Office, Washington, D.C. Volchok H. and de Planque G. (Eds) (1982) EML Procedures Manual, HASL-300,25th edn. Department of Energy, New York. Walker C. R., Short B. W. and Spring H. S. (1980) In Radioelement Analysis: Progress and Problems (Ed. Lyon W. C.) p. 101. Ann Arbor Science Publishers.
1987 Copyright
6
0883-2889/87 $3.00 + 0.00 1987 Pergamon Journals Ltd
Measurement of 99Tc in Low-Level Radioactive Waste from Reactors using 99mT~as a Tracer J. E. MARTIN* School
of Public
Health,
University
and
of Michigan,
J. M. HYLKO Ann Arbor, MI 48109-2029, U.S.A
(Received 20 October 1986; in revised form20 November 1986) waste procedure has been developed for measuring 99Tc in major low-level radioactive streams from commercial nuclear power facilities which contain numerous potential contaminants. Separation was by alkaline precipitation and anion exchange chromatography; trace amounts of “Co and other radioactive contaminants were removed by solvent separation. Pure separations were obtained with average radiochemical yield of 47.3 5 4.3% determined by a 99mTctracer. The lower limit of detection for
A radiochemical
%Tc was found to be 0.16 pCi/g based on IO-g sample and a IOO-min counting time.
Introduction Technetium-99 is an important constituent of low-level radioactive waste (LLW) and has received particular attention as a source term for shallow land burial sites (U.S. NRC, 1982). Regulations promulgated by Title 10, Part 61 in the Code of Federal Regulations (10 CFR 61) require concentrations of p9Tc to be reported on LLW shipping manifests so site inventories can be established. This requirement is basic to determining the potential impact of any new sites operated by regional LLW compacts in accordance with the Low-Level Waste Policy Act of 1980 (U.S. Congress, 1980). Technetium-99 has a half life of 2.13 x 10’~ (Kocher, 1981); however, Coursey et al., (1984) in recent studies suggest that 2.111 + 0.012 x 10’~ may be a more appropriate value. The nuclide is a pure beta-particle emitter with a maximum beta energy of 292 keV. Its primary mechanism of production is by thermal neutron fission of 235U and 239Pu in nuclear reactors; thus, nuclear power plants represent a significant source of 9qc. Production of 9qc also occurs at a low rate by neutron activation of stable 98M~ which exists in fuel as a stable fission product or outside fuel as a trace constituent in reactor structures (Till et al., 1985). Various analytical procedures have been used to measure ‘%I’c in samples. Chu and Feldstein (1984) list several such methods and their limitations for measuring 99Tc in environmental samples using betaemitting Tc tracers to measure chemical recovery. Walker et al. (1980) report the use of liquid scintillation counting as a quick and inexpensive technique for measuring !@Tc if chemiluminescence is carefully * Author
for correspondence.
eliminated, and Coursey et al. (1984) have used it as a means of standardizing 99Tc solutions. Foti et al. (1972) have used neutron activation analysis to measure 99Tc at the picocurie level. The induced ‘“Tc activity was measured by y-ray spectroscopy immediately after irradiation. Since the half life of the lWTc is only 16 s, computerized data acquisition and analysis are required to process the y-ray spectra. Technetium can also be separated by distillation and solvent extraction of aqueous-organic phases. The principal disadvantage of this method is the introduction of acid or organic mixtures to reduced the pertechnetate (TcO;) ion which causes difficulties in subsequent separation steps using ion-exchange chromatography (Anders, 1960). The feasibility of 99Tc analysis by gravimetric, spectrophotometric or polarographic techniques has not been demonstrated since these methods are not sufficiently sensitive for environmental samples (Till et al., 1985). In general, a reliable procedure for measuring 9?c in the diversity of LLW samples from reactors has not been reported. Since the primary source of 99Tc in LLW in nuclear reactor waste streams, it is important that the method be applicable to quantitating the concentrations of 9qc in the various liquid and solid media such wastes comprise. Because such waste streams contain numerous beta emitters that could interfere with the accuracy of 99Tc measurements, it is important to assume that the measured concentrations represent the 9‘Tc and not a contaminant artifact; thus, pure separations are required. The objectives of this study were to develop such a procedure, to establish a lower limit of detection for it, and to determine the value of using 99Tc as a tracer to measure chemical recovery. Although these studies did not consider 9qc in other non-fuel-cycle LLW, such as that from research and medical institutions, 447
J. E. MARTIN and
448
the same methodology and approach be applicable for these wastes.
are believed to
Method Samples of LLW types representing about 6 months operation were obtained from the Big Rock Point and Palisades Nuclear Plants in Michigan. Detailed summaries of the types of samples analysed have been reported for 14Cfor these plants and others by Martin (1985). Other data on 9’Tc are presented by Hylko (1986) on studies of the effect of wall absorption in polyethylene containers on measurements of 99Tc in liquid waste samples. These studies showed that about 50% of the 99Tc absorbed onto the container, but that this could virtually be eliminated by sample pretreatment with paper pulp and HCI. This sampling feature is incorported into the analytical method now used for 99Tc determinations in LLW. The radiochemical procedure reported by Chu et al. (1984) was modified for separation of 99Tc from mixed activation, corrosion, and fission products. Liquid samples containing paper pulp for pretreatment were filtered through a 0.2 pm filter to remove particulate matter so the filtrate and filterable materal could be analyzed separately. A measured volume of 99mTc extracted from a 99Mo-99mTc generator was added prior to the chemical separation as a tracer to determine chemical recovery. A 2 x 2-in. NaI(TI) crystal was used for counting the 99mTc tracer; it was not necessary to standardize the tracer since a ratio method was used on identical volumes added to each sample and counting before and after separation. The activity of the tracer spikes was low enough to minimize the effect of dead-time losses. Solid samples were dissolved under reflux conditions in 7.5 M HNO,. If solid material remained, the sample was filtered and the solid material was refluxed a second time and analyzed separately. Five milliliters of Sr” and 5 mL of CO’+ were added as carriers; each contained a nitrate concentration of 20 mg/mL in solution. The samples were then treated with NaOH and Na,CO, to precipitate the actinides, cesium, strontium, and the transition metals leaving 99Tc in solution. The filtrate was acidified by adding 7.5 M HNO, to the solution until the orange-colored methyl red endpoint was achieved, then 12 mL more of the acid was added which changed the solution color to a clear pink. The solution was heated gently to remove the CO, produced by the acidification of the carbonate, and the solution was diluted with deionized water to 870 mL yielding a final solution containing about 0.1 M nitric acid. The 99Tc solution was concentrated on a Dowex 1 x 4 anion exchange resin further separating it from other trace nuclides present in the weakly acidic solution. The resin was washed with 500 mL of 0.1 M HNO, and the concentrated 99Tc was eluted from the resin at full flow with 100 mL of 5 M HNO, at a
J. M.
HYLKO
recovery of about 90%. Evaporation temperatures were held between 80 and 90 C for the column elutions to reduce sample loss. which was less than 10% under such conditions. The eluate pH was increased to between 10 and 11 using concentrated NH,OH, followed by addition of 10mL of 0.5 M tetraphenylarsonium chloride. The mixture was transferred to a separatory funnel containing 100 mL of chloroform and was shaken for 2 min. The chloroform fraction was added to a second separatory funnel containing 50 mL of 0. I M NH,OH and was shaken for 2min. The NH,OH fraction was saved and the alkaline wash was repeated. The solution was evaporated to less than 10mL and transferred to a ringed stainless-steel planchct and evaporated to dryness. A low-background gas-flow proportional counter was used to measure the separated samples. The counter was calibrated for 99Tc radioassay using a solution of “Tc standard (Amersham Corp., Arlington Heights, Ill.) evaporated on ringed stainless steel planchets. An identical volume of unprocessed 99mTc spike (100% recovery) was counted followed by a processed samples with a Nal(T1) crystal and a single channel analyzer for chemical recovery. Total counting time was varied to reduce counting error to less than 5%. The net counting rate of the 99mT~in each sample was corrected for decay to the time the unprocessed spike was counted in order to determine chemical recovery. The samples were stored for a minimum of 60 h for 99mTc to decay before counting for 99Tc. The 99Tc activity per gram of sample, A (99T~), was calculated by: A (“Tc) =
C net EWY
2.22 x IO’* d/mCi
where C,,, is the net beta count rate (cpm) of the sample with background subtracted (based on a counting time of lOOmin), E is the 99Tc counting efficiency, W is sample weight, and Y is the chemical yield based on the decay-corrected count rate of the y9Tc in the processed sample divided by the count rate of the 99mTc spike (100% recovery). A lower limit of detectability based on the method of Currie (1968) was determined.
Results and Discussion The experimental procedure used was orders of magnitude more sensitive than the level of 0.03 Ci/m’ required by 10 CFR 61 for LLW waste characterization (U.S. NRC, 1982). Considerable work was done to establish the procedure and verify its yield, efficiency, and reproducibility. Table 1 contains the results of analyses made with 99Tc and 99mT~tracers to establish recovery, which was found to be 47.3 + 4.3% for liquid samples through the precipitation, anion-exchange, and solvent separation procedure; solid sample recoveries were slightly better. Memory checks were run following each processed standard
%Tc in reactor Table
that complete equilibrium of the 99mTctracer and 9qc in the samples was accomplished independent of the sample matrix. This was confirmed by analyzing two sets of identical samples.
1. Chemical recovery for Drocessed samoles Sample
Recovery (%)
Reactor water (treated) Reactor water (untreated) Concentrates Resins Crud Waste filter Smears
45.01 47.8’ 46.5’ 48.57 48.5t 56.0 45.0
449
wastes
Sample con taminan ts
l Average of recoveries for filtrate and the filter. able fractions for two samples. t Average of two samples.
and in the course of the experiments to establish that residual deposition in the apparatus was indeed insignificant. Background counts with blanks were about 1.4-1.6 cpm, which with a lOO-min counting time yielded a minimum detectable activity (Currie, 1968) of about 0.16 pCi/g for a 10 g sample at 47.3 k 4.3% recovery. Solid samples The method was applied to solid samples by refluxing the sample in 7.5 M nitric acid. This refluxing procedure proved to be efficient for uniform dissolution of the samples. Solid material that did not dissolve during initial pretreatment was analyzed a second time; however, sample activity was always equivalent to background levels. When excess refluxing was necessary for dissolving such samples, some chemical recoveries dropped below SO%, demonstrating the value of the 99mT~tracer for non-liquid samples. Identical samples analyzed a second time from the same or similar containers were, with 95% confidence, statistically equal to the initial 99Tc activity that was measured. This strongly suggests
Because LLW samples contain significant radioactivity from numerous nuclides, the presence of contaminants in the %Tc samples is always possible even after chemical separation. Cobalt-60 is the most common nuclide found after radiochemical separations, probably because it represents the highest measured concentration in most LLW samples and because of its chemical properties. Decontamination of 99Tc from trace amounts of ‘j”Co and other contaminants was achieved by the solvent separation step. In order to establish the 99Tc was in fact separated from higher energy beta-emitters, a HarleyHallden plot (Volchok and de Planque, 1982) was determined for each sample. The slopes of the plots were determined to be statistically equal to unity at the 95% confidence level (Hylko, 1986). Liquid scintillation counting was also used to determine the presence of other beta emitting contaminants. Figure 1 shows a plot of the beta spectra of a 99Tc standard and two LLW samples (without quench correction) determined with a Beckman LS 6800 Liquid Scintillation Analyzer (Beckman Inc., Irvine, Calif.) which has a logarithmic amplifier. The spectra of the two LLW samples show comparable energy distributions with that of the 9qc standard, providing additional confirmation of the sample purity. Since 6oCo contamination (maximum beta energy of 0.317 MeV) is a likely contaminant that could by masked by 99Tc in liquid scintillation counting, the samples were also gamma counted using an
l.O-
‘.._4-
.... .il
-0
I&
260
300 CHANNEL
Fig. 1. Liquid
scintillation
460
So0
“To
Reactor
Concentrates (H+213)
SbO
NUMBER
analyzer (Beckman Model LS 6800) beta spectra of two ‘Vc separations reactor waste samples compared to a WTc standard.
of
J. E. MARTINand J. M. HYLKO
450
intrinsic germanium detector on the presumption that beta-emitting contaminants in LLW would also emit y rays. Such contaminants would distort the proportional counter results and liquid scintillation analyses due to the spectral distribution of beta energies. All beta and gamma spectra of processed LLW samples indicated that the samples were free of radionuclide contamination and were equivalent to processed 99Tc standards.
Summary and Conclusions This study was undertaken to develop a methodology for measuring 99Tc in various liquid, solid, and semi-solid low level radioactive waste samples. A modification of Chu’s method (Chu and Feldstein, 1984) using a combination of alkaline carbonate precipitation, anion-exchange chromatography, and solvent separation provided excellent isolation of 9’?c from mixed activation, corrosion, and fission products for the usual forms of reactor LLW that require classification. Good separation of qc was obtained with yields of about 47.3 f 4.3% and the method eliminated any significant interference by the myriad of other constituents in the LLW samples. Technetium-99m, the pure gamma-emitting isomer of ‘?c, was found to be an excellent tracer for determining chemical recovery because its short half life (6.007 h) allowed complete decay prior to counting for 9qc. Low background beta counting of processed samples allowed reliable detection of ?c concentrations in LLW samples of 0.16 pCi/g based as a 10 g sample and 47.3% chemical recovery.
References Anders E. (1960) Radiochemistry qf Technetium, NASNS3021 National Academy of Sciences-National Research Council, Washington D.C. Chu N. Y. and Feldstein J. (1984) Radiochemical determination of technetium-99. Talanta 31, 809. Coursey B. M., Gibson J. A. B., Heitzmann M. W. and Leak J. C. (1984) Standardization of technetium-99 by liquid scintillation counting. ht. J. Appl. Radial. Isot. 35, 1103. Currie L. A. (1968) Limits for qualitative and quantitative determination. Anal. Chem. 40, 586. Foti S., Delucchi E. and Akamian V. (1972) Determination of picocurie amounts of WTc by neutron activation. Anal. Chim. Acta 60, 261. Hylko J. M. (1986) The Analysis of %Tc and “‘CS in Low -Level Waste Samples from- Commercial Nuclear Power Facilities. Master’s Thesis (Unuublished). School of Public Health, University of Michigan, Ann Arbor, MI 48109, U.S.A. Kocher D. C. (1981) Radioactive Decay Data Tables. DOG/TIC 11026, 108. Martin J. E. (1985) Carbon-14 in low-level radioactive waste from two nuclear power plants. Health Phys. 50, 57. U.S. Nuclear Regulatory Commission (1982) Licensing requirements for land disposal of radioactive waste. Federal Re.&ter 47, 51446 (27 December 1982). Till J. E., Sh& R. W. and Hoffman F. 0. (1985) Uranium Fuel Cvcle-A Review of Data for Technetium. NUREGI CR-3738. U.S. Nuclear- Regulatory Commission, Washington, D.C. U.S. Congress (1980) Low-Level Waste Policy Act qf 1980, Public Law 96-573, 94 Stat. 3347. U.S. Government Printing Office, Washington, D.C. Volchok H. and de Planque G. (Eds) (1982) EML Procedures Manual, HASL-300,25th edn. Department of Energy, New York. Walker C. R., Short B. W. and Spring H. S. (1980) In Radioelement Analysis: Progress and Problems (Ed. Lyon W. C.) p. 101. Ann Arbor Science Publishers.