Article pubs.acs.org/ac
Liquid Scintillation Counting Methodology for Remedy for Radiopharmaceutical Waste
99
Tc Analysis: A
Mumtaz Khan† and Wooyong Um*,†,‡ †
Division of Advanced Nuclear Engineering, Nuclear Engineering Laboratory, Engineering Building I, Pohang University of Science and Technology, 77 Cheongam-ro, Nam-Gu, Pohang, Gyeongbuk 790-784, Republic of Korea ‡ Pacific Northwest National Laboratory, 902 Battelle Boulevard, Richland, Washington 99354, United States ABSTRACT: This paper presents a new approach for liquid scintillation counting (LSC) analysis of single-radionuclide samples containing appreciable organic or inorganic quench. This work offers better analytical results than existing LSC methods for technetium-99 (99gTc) analysis with significant savings in analysis cost and time. The method was developed to quantify 99gTc in environmental liquid and urine samples using LSC. Method efficiency was measured in the presence of 1.9 to 11 900 ppm total dissolved solids. The resultant quench curve proved to be effective for quantifying spiked 99gTc activity in deionized water, tap water, groundwater, seawater, and urine samples. Counting efficiency was found to be 91.66% for Ultima Gold LLT (ULG-LLT) and Ultima Gold (ULG). Relative error in spiked 99gTc samples was ±3.98% in ULG and ULG-LLT cocktails. Minimum detectable activity was determined to be 25.3 and 22.7 mBq for ULG-LLT and ULG cocktails, respectively. A preconcentration factor of 1000 was achieved at 100 °C for 100% chemical recovery.
T
echnetium-99g (99gTc) is a challenging radionuclide from the perspective of radioactive waste management. It presents many problems related to safe disposal because of its high mobility and volatility in its environmentally stable form of pertechnetate (99gTcO4−). Pertechnetate changes speciation in oxic and anoxic environments. The high water solubility of 99g TcO4−, its long half-life of 2.13 × 105 years, long residence time in the ocean, and high plant uptake behavior make it a new oceanographic tracer for investigation of water movement. Many studies of 99gTc disposal, chemical characterization, and water body movement have been previously conducted but still require an accurate analysis of 99gTc in its various chemical forms. Its chemical form affects analysis results, especially in liquid scintillation-type of radiochemical analysis.1 Technetium-99m (99mTc) is used in the medical field for diagnosis purposes with a large number of organic carriers. These include 99mTc-mercaptoactyltriglycine,2 99mTc-gluconate,3 99mTc-dimercaptosuccinic acid,4,5 99mTc-methylene-diphosphonate, 99mTc-dicarboxypropane-diphosphonate, 99mTcethylenediaminetetramethylene-phosphonate,6 and 99mTc-pingyangmycin.7 99mTc is used in 80% of diagnostic nuclear medicine procedures. Almost 30 million examinations are conducted worldwide using this isotope, because 99mTc has a half-life of 6 h and decays to 99gTc. Monitoring and safe disposal of 99gTc from human urine is very important, and concern is increasing every day as global use of 99mTc has increased by more than 4.5 × 1014 Bq per week and is increasing continuously. In 2020, demand is expected to increase 32% as compared to current market demand. Patients’ excretions are discharged to sewer systems without quantification of 99gTc due to lack of a rapid procedure.8,9 Most of the existing 99gTc analysis procedures10−13 are based on purification © 2015 American Chemical Society
of the sample matrix, which requires a trained analytical chemist in a hospital or nearby analytical laboratory. Keeping in view all the practical problems of hospitals and laboratories, a simple and direct analytical method needs to be developed for quantification of 99gTc in urine as well as municipal waste samples. Previous methods have high minimum detectable activity (MDA) and low figure of merit (FOM).10−15 The method presented here has a low MDA and better sensitivity as compared to previous methods. Our procedure also addresses the interference of organic matter on the quantification of 99gTc activity. This method is simple and fast, and the quenchcorrected curve gives accurate results.16−19 Generally, purification and isolation of a radionuclide from the sample matrix is a prerequisite before counting radioactivity by LSC. Presence of different ions or species, even if they are nonradioactive, can change the LSC measured activity of the desired radionuclide due to their quenching capability. Most of the instruments used to quantify 99gTc are sensitive to the dissolved solids present in the sample, such as inductively coupled plasma (ICP) mass spectroscopy (MS), ICP optical emission spectrometry (OES), and ion-chromatography (IC). However, the processes for purification and isolation of radionuclides are also time, cost, and labor-intensive for both normal and emergency measurements. A novel feature of the current method is that it does not require isolation of 99gTc from an aqueous matrix,12 as the effect of any nonradioactive salt or organic impurity has been incorporated via quench curve efficiency tracing. Received: June 17, 2015 Accepted: August 13, 2015 Published: August 13, 2015 9054
DOI: 10.1021/acs.analchem.5b02279 Anal. Chem. 2015, 87, 9054−9060
Article
Analytical Chemistry Table 1. Method-Optimized Values for Two Different Cocktails B (cpm) ULG-LLT
ULG
default 20 B (cpm) default 22
optimized 10
optimized 8
ε
MDA (mBq) default optimized 45.58 25.27 MDA (mBq) default 50.12
optimized 22.69
FOM
default 101.66
optimized 91.66
default 516.81
ε default 102.22
optimized 840.27 FOM
optimized 91.66
default 474.95
optimized 1050.34
Table 2. Optimized Values and Studied Ranges of LSC Parameters for Both Cocktails name of parameter
ULG
ULG-LLT
study range
vial type delay before burst (ns) count mode counting window (KeV) counting time (minutes) precount delay (minutes) external standard cocktail volume (mL) coincidence time (ns) pH of sample and cocktail mixture sample + cocktail mixture temp (°C)
glass 200 high sensitivity 10−160 120 1 0.5 2S% 10 18 4.5−5.5 25 °C
glass 200 high sensitivity 10−160 120 1 0.5 2S% 15 18 4.5−5.5 25 °C
glass/polyethylene 80−800 normal/sensitive 0−293 10−180 1−180 5−20 10−30 -
in DIW. Radioactive solutions of 99gTc (1 mL) sample having activity 1 Bq/mL were mixed with 15 mL of ULG-LLT cocktail in 20 mL volume glass and polyethylene vials, separately. Blank solutions were also prepared using 1 mL of DIW and 15 mL of ULG-LLT in both glass and polyethylene vials. Using ULGLLT standard and blank preparation procedures, one 99gTc standard and one DIW blank solution were prepared with 10 mL of ULG cocktail in glass vials. ULG cocktail was not used at volumes greater than 10 mL because more than 10 mL of ULG produces a high background. Both of these standard solutions were used to optimize the LSC counting methodology for all LSC parameters. Method Validation. Five different types of samples were prepared. Tap water (TW) was collected from the laboratory. Seawater (SW) was collected from the Pohang seashore in the East Sea. Ground water (GW) was collected from the Wolsung radioactive waste repository site. DIW was also collected from the laboratory deionized water distillation unit. Urine sample was collected from a volunteer worker at our laboratory. All samples were collected in 2 L polypropylene bottles. The bottles were stored in a refrigerator at 4 °C before analysis. On analysis day, each sample was filtered through a 0.45 μm polyvinylidene fluoride (PVDF) Whatman syringe filter. One milliliter of each filtered sample was spiked with 1 mL of 99gTc (1 Bq) and mixed with 10 or 15 mL of ULG and ULG-LLT cocktails, respectively. For blank sample preparation, 1 mL of each sample was spiked with 1 mL DIW without 99gTc and mixed with 10 or 15 mL of ULG or ULG-LLT cocktail, respectively. All samples were prepared under dark conditions. River Sample. In addition to the five sets of samples mentioned above, a water sample of 1000 mL was collected from the nearby Hyeongsan River in a polypropylene container for environmental sample testing. It was filtered through Whatman filter paper (47 mm) and 0.45 μm polyvinylidene fluoride (PVDF) Whatman syringe filter sequentially. Then, it was spiked with 1 mL of 99gTc (1 Bq) followed by preconcentration at 100 °C, and sample volume was reduced to 1 mL. After cooling, the concentrated river water was placed
A liquid scintillation counter (LSC) is a simple instrument, but the measuring procedure involves many complicated parameters that should be considered for accurate analysis of actual samples. Keeping in view all the technology constraints, this method optimizes all the instrument variables that control the analysis results.20−23 The need to accurately measure the quantity of 99gTc activity, released from hospitals to the sewer system is growing because of public and sewer workers’ radiation safety and concerns. Hospital’s off streams, where sanitary workers are involved in the management of waste, is necessary to make sure of personnel and environmental safety. The objective of this research is to compare radiation detection sensitivity of different cocktails for different types of common cosolutes in order to measure accurate 99gTc activity in “real world” matrices. In this method, background (B), MDA, efficiency (ε), and FOM of various cocktails have been studied. However, we focused on “FOM” as a better parameter for describing sensitivity rather than only “efficiency” of the radiation detection.
■
MATERIALS AND METHODS Materials and Instrumentation. All the reagents used were analytical grade. ULG-LLT and ULG cocktails were purchased from PerkinElmer. Nitromethane (99+% purity) was purchased from Across Organics, U.S.A.. Technetium standard (99+% purity) was purchased from Eckert & Ziegler Isotope Products, California. Deionized water (DIW) was prepared using a Direct-Q 5 UV distillation unit. A PerkinElmer Tri-Carb 2910TR LSC equipped with a Barium-133 (133Ba) external standard was used for counting samples at laboratory temperature of 25 °C. Total dissolved solids (TDS) were calculated from electrical conductivity measurements in spiked samples using a Thermo Scientific Orion Star A211 multimeter. Standards Preparation. Method Development. Technetium-99 (which was in chemical form of ammonium pertechnetate, NH4TcO4) stock solution of 10 Bq/mL was prepared in DIW from a 99gTc ampule having activity of 185 000 Bq/mL. A working solution of 1 Bq/mL was prepared 9055
DOI: 10.1021/acs.analchem.5b02279 Anal. Chem. 2015, 87, 9054−9060
Article
Analytical Chemistry Table 3. Comparison of Different LSC Methods ε (%)/FOM
MDA (mBq)
cocktail
counting time
analyses time
matrix
90
17
Ultima Gold AB
2h
1 week
biota/sediments
97.5
100
Opti-phase Hisafe
22.69
Ultima Gold/Ultima Gold LLT
not reported, most probably higher than 2 h 1−2 h
urine
91.66/1050.34
30 min with 12 h precount delay 2h
in a 20 mL glass vial, and 10 mL of ULG was added. It was counted under the desired protocol. Quench Standards. One set of nine glass vials was prepared with 1 mL 99gTc (1 Bq) and 15 mL of ULG-LLT cocktail. Before adding the quench indicator, each vial was counted in the LSC for 120 min to verify that 99gTc was preset. All vials should have the same amount of radioactivity. No vial was found having more than 2% deviation from the mean value. Later on, variable quantities of nitromethane (0, 10, 15, 20, 25, 30, 35, 40, or 45 μL), a known quenching agent, were added to the nine vials. Another set of nine quench standards was prepared with 10 mL of ULG cocktail. A procedure similar to that used for ULGLLT was repeated for ULG. Vials were placed in the dark, inside the LSC, before starting the experiment. The most highly quenched vial was placed at position 1, and the unquenched vial was placed in the last position within the cassette.
sea, river, tap, ground water/urine
instrument Wallac 1220 Quantulus Tricarb 2810 TR Tricarb 2910 TR
ref 10 11 present work
Figure 3. Effect of coincidence time on method sensitivity.
thing is to select the right choice of scintillator and its quantity. Careful selection of the cocktail and its volume used influences the radionuclide measurement efficiency.20 Table 1 shows that the type of cocktail has a significant effect on the analysis results of the LSC radiometric method. The most important variables governing the superiority of any counting method are B, MDA, ε, and FOM. Both B and MDA should be as low as possible, whereas ε and FOM should be as high as possible. The values for optimized and default variables for both types of cocktails are shown in Table 1. In addition, Table 1 shows that both cocktails have equal ε values, but ULG has lower B, which increases its FOM. Low B value has also reduced the limit of detection for ULG as compared to ULG-LLT, which shows ULG has more resistance to cosmic rays. However, both cocktails have lower B and better FOM in comparison with the default parameter method of the LSC 2910TR. Table 1 clearly demonstrates that ULG is a better cocktail for quench-free 99gTc samples. LSC Parameters. Every cocktail has its own decay constant; when it absorbs energy from a β particle, it emits radiation, which may be slow or fast. This emission of light should be compatible with instrument parameters such as delay before burst (DBB) and coincidence time. All the instrumentoptimized parameters and the studied range are shown in Table 2 for both types of cocktails. Table 2 also lists physical factors that affect the yield of light and method sensitivity. Comparison with Previous Studies. Technetium-99g can be measured by many analytical techniques including resonance ionization MS, accelerator MS, ICP−sector-field MS, ICP-MS, ICP-quadrupole MS, Geiger-Müller detector, and LSC.12 Each method has its own advantages and drawbacks, and the type of analysis often depends on the instrument available. Our method was developed to measure 99gTc by LSC. Table 3 compares our method with other published reports that describe 99gTc analysis. Only two references were selected as these are recently published and most relevant.10,11 The MDA of the method reported by Wigley et al.10 is slightly lower than our method. However, our method is better in terms of
■
RESULTS AND DISCUSSION Comparison of Cocktails. Among the factors which control the development of the radiometric method, the first
Figure 1. Counting uncertainty with time.
Figure 2. Senstivity-controlling parameters of LSC.
9056
DOI: 10.1021/acs.analchem.5b02279 Anal. Chem. 2015, 87, 9054−9060
Article
Analytical Chemistry
Figure 4. Impact of organic quench on method sensitivity.
Figure 5. Effect of TDS on cocktail behavior.
time period because our method has relatively short sample preparation times. For our method, it is not required to isolate 99g Tc from the sample matrix, whereas the Villar et al.11 method needs an automatic isolation assembly which increases the cost of the analysis as well.
efficiency, variety of cocktails, much short analysis time, variety of sample matrices, less-expensive counting instrument, and emergency analysis. Our method is also better than the method reported by Villar et al.11 in terms of low MDA, variety of cocktails, no precount delay as well as range of sample matrix. Using our method, many samples can be analyzed in a short 9057
DOI: 10.1021/acs.analchem.5b02279 Anal. Chem. 2015, 87, 9054−9060
Article
Analytical Chemistry
distinguished from true beta events because they are characterized by a series of low amplitude after-pulses that follow the initial prompt pulse. True β scintillation pulses have fewer after-pulses associated with them. Note that background pulses have more prominent after-pulse characteristics, which extend over time. The time-resolved LSC is designed to reduce background by evaluating each event for the presence of these after-pulses, and when a series of after-pulses is detected, the time-resolved LSC characterizes the event as background and rejects it.22 The Model 2910TR LSC is equipped with only normal and high sensitivity mode. On applying “high sensitivity” mode after proper selection of the energy window, background was reduced to 8 counts per minute (CPM), whereas FOM was increased to 822.37, as shown in Figure 2. Delay Before Burst. Delay before burst (DBB), a feature in LSC optimizes background that is generated according to different delay time’s intervals. After-pulses of quenchable and unquenchable background may persist up to 900 ns and 5 × 106 ns, respectively. The 2910TR LSC is equipped with a DBB range of 75−800 ns. Our desired value was optimized at 200 ns. The optimized DBB increases B from 8 to 10 CPM, but FOM was also increased from 822.37 to 850.49, as shown in Figure 2. This shows that the delayed portion of the “real beta event” was incorporated more than the background counts portion.21 Coincidence Circuit. Coincidence circuit (CC) time was adjusted, as shown in Figure 3. The CC time tells whether the signal output from the photomultiplier tubes reaches the coincidence circuit simultaneously. It is necessary to fix its value to avoid any false thermal noise signal.1 Figure 3 shows that there is no significant change in FOM, and no further change is possible after changing the coincidence circuit time from 18 to 30 ns. The sharp reduction in FOM before 15 ns indicates that the instrument will lose ε at or below 15 ns. Therefore, 18 ns was the selected operating CC time value, where FOM was maximized and B was stable. The sharp reduction in B and FOM below 15 ns also shows that B is reduced at the cost of real beta counts in that range. Physical Factors. Physical factors include sample container material, pH, volume of the cocktail, and temperature of the solution. All of these factors were also considered. Optimized values for ULG and ULG-LLT cocktails are given in Table 2. Quench Variation. The presence of any impurity (chemical or color substances and solution opacity) severely affects the yield of light and leads to incorrect quantification of radioactivity.1 Effects of both inorganic and organic impurities
Table 4. Comparison of Different Organic and Inorganic Matter sample
TDS (ppm)
tSIE (UGLLT)
% ε (ULGLLT)
tSIE (ULG)
%ε (ULG)
DIW urine
1.9 4220
447.32 428.62
91.66 88.3
488.11 457.9
91.66 88.33
LSC Factors. Many factors that control the accuracy of results are related to the instrument, such as DBB, coincidence circuit (CC) timing, luminescence control, time-resolve mode, energy window level setting, and statistical accuracy controlling parameters. Among these factors, the dominant factors that lower B and raise FOM are explained in the following sections. Counting Time. Counting uncertainty is represented by 2S %, which is a very important parameter with respect to activity counting. Accuracy in activity counting is highly dependent on the counting time, as shown in Figure 1, which also indicates that there is stability in LSC counts after 120 min of counting time. All measurements were performed for 120 min, which has the lowest counting uncertainty at 95% confidence level. Energy Window. Default values of B and FOM (= ε2/B) in Figure 2 were calculated for the full energy range of the window at 75 DBB. Later on, the energy window was set at the 10−160 keV range instead of 0−2000 keV. The energy range was narrowed to avoid the effect of high energy cosmic rays which penetrate the shielding and produce counts in the cocktail. The cosmic ray contribution is called quenchable background. This portion of background usually contributes in the higher energy end, above 160 keV, and was distinguished by cutting off the energy level at 160 keV. The lower energy end
