Applied Radiation and Isotopes 87 (2014) 282–286
Contents lists available at ScienceDirect
Applied Radiation and Isotopes journal homepage: www.elsevier.com/locate/apradiso
Micelle size effect on Fe-55 liquid scintillation efficiency Denis E. Bergeron n, Lizbeth Laureano-Pérez Physical Measurements Laboratory, National Institute of Standards and Technology, 100 Bureau Dr., Gaithersburg, MD 20899-8462, USA
H I G H L I G H T S
No efficiency reductions attributable to the micelle size effect were observed. Calculated micelle corrections are small compared to other uncertainties. Metal ions did not affect micelle size or fluorescence quenching. Tracing 55Fe efficiencies with 54Mn reduced model dependence.
art ic l e i nf o
a b s t r a c t
Available online 1 December 2013
We used efficiency tracing techniques to study the micelle size effect on liquid scintillation counting of the electron capture nuclide, 55Fe. We determined micelle hydrodynamic diameters for specific LS cocktails via dynamic light scattering, and sought trends in efficiencies as a function of micelle size. The presence of Fe3 þ or Mn2 þ ions in the cocktails did not significantly affect micelle sizes or fluorescence quenching. We did not detect any reductions in counting efficiencies due to the micelle size effect. Published by Elsevier Ltd.
Keywords: Reverse micelle Liquid scintillation counting Cocktail Mn-54 H-3
1. Introduction We recently undertook a series of dynamic light scattering (DLS) measurements to determine the size of micelles in several commercial liquid scintillation (LS) cocktails (Bergeron, 2012). An electron emitted from a radionuclide residing within a reverse micelle will lose energy to the aqueous medium before interacting with the scintillant residing in the organic phase, resulting in a reduction in counting efficiency typically referred to as “the micelle effect” or “micelle size effect” (Grau Carles, 2007; Kossert and Grau Carles, 2008, 2010). Our measurements revealed that micelles in commercial LS cocktails are smaller than previously assumed (based on measurements by Rodríguez et al., 1998), so that the micelle size effect is smaller than was thought. We showed that even for the electron capture nuclide 55Fe, the effect would reduce counting efficiencies by o0.2% under typical counting conditions (Bergeron, 2012). In this work, we set out to determine whether it might be possible to observe the micelle size effect under ordinary experimental conditions. We selected 55Fe because, as has been described previously (Grau Carles, 2006, 2007; Kossert and Grau Carles, 2010; Bergeron, 2012), its low energy Auger electron emissions make it particularly sensitive to the effect. Through careful cocktail selection, we sought to maximize the achievable range of micelle sizes. We were also careful to select cocktails with ionic and nonionic surfactants since the
n
Corresponding author. Tel.: þ 1 301 975 2282; fax: þ1 301 926 7416. E-mail addresses: [email protected], [email protected] (D.E. Bergeron).
0969-8043/$ - see front matter Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.apradiso.2013.11.080
location of the nuclide within the micelle, which may be affected by the surfactant type (Pileni et al., 1985; Pant et al., 1998; Andrade and Costa, 2002; Faeder and Ladanyi, 2000, 2001, 2005; Stahla et al., 2008), should affect the magnitude of the efficiency reduction. Since the variation of micelle size is achieved by varying the total aqueous fraction in the cocktail, and since the addition of water changes the general quenching properties of the cocktail, the experimental scheme included CIEMAT/NIST efficiency tracing (CNET) with both 3H and 54 Mn. It was hoped that careful efficiency tracing would help to minimize contributions from quenching mechanisms that might obscure the detection of the micelle size effect.
2. Methods 2.1. Sample preparation Two experiments were performed. All of the selected scintillants were diisopropyl naphthalene (DIN) based. Table 1 provides details on the scintillation cocktails used in both experiments. In Experiment 1, cocktails were prepared from HiSafe II and HiSafe III (PerkinElmer, Waltham, MA, USA).1 The scintillants were 1 Certain commercial equipment, instruments, or materials are identified in this paper to foster understanding. Such identification does not imply recommendation by the National Institute of Standards and Technology, nor does it imply that the materials or equipment identified are necessarily the best available for the purpose.
D.E. Bergeron, L. Laureano-Pérez / Applied Radiation and Isotopes 87 (2014) 282–286
283
Table 1 Summary of the LS cocktails used in the experiments. Each series includes NS separate samples, each with a different total aqueous fraction, f, covering the indicated range. CFe and CMn indicate the total concentration of the metal ions; the metal ions are introduced as carrier solution, or (indicated by italicization) standard solutions of the appropriate radionuclide. For more details on the surfactants and other components of the scintillants, refer to the manufacturer's specifications and material safety data sheets. Radionuclide solution Experiment 1 Fe H Blank 55 Fe 3 H Blank Experiment 2 55 Fe 54 Mn 3 H Blank 55 Fe 54 Mn 3 H Blank 55
Scintillant
Surfactant information
Ns
CFe/(μg g 1)
HiSafe II
In addition to nonionic surfactants, includes 10–20% sodium dioctyl sulphosuccinate (anionic surfactant)
HiSafe III
20% to 40% ethoxylate polymers (nonionic surfactants)
6 6 8 5 5 7
0.065 0.065 0 or 0.065 0.065 0.065 0 or 0.065
Ultima Gold
In addition to nonionic surfactants, indludes r 2.5% sodium dioctyl sulphosuccinate (anionic surfactant)
HiSafe III
20–40% Ethoxylate polymers (nonionic surfactants)
6 6 8 7 5 5 7 6
0.67 0.67 0 or 0.67 0 or 0.67 0.67 0.67 0 or 0.67 0 or 0.67
3
selected to include one formulation that contains an ionic surfactant, and one that includes only non-ionic surfactants (Table 1). Matched 55Fe, 3H, and blank cocktails were prepared with a range of aqueous fractions (f¼ vw/vtot; where vw is the total aqueous sample volume and vtot is the total cocktail volume, including added aqueous sample) selected to maximize the range of micelle sizes as measured by DLS in terms of hydrodynamic diameter (Bergeron, 2012). The 55Fe cocktails were prepared gravimetrically from a 14.2 kBq g-1 55Fe solution with 19 μg g 1 of Fe3 þ (as FeCl3) in 1 mol L 1 HCl (NIST, 2006). The matched 3H (NIST, 2008) cocktails and blanks contained Fe3 þ carrier solution. While the Fe3 þ content for all cocktails was the same, the desired range of f was achieved via addition of different amounts of distilled water, so that the cocktails with highest f had the lowest Fe3 þ concentration. In Experiment 2, HiSafe II was replaced by Ultima Gold (PerkinElmer, Waltham, MA, USA). All cocktails contained the same total amount of Fe3 þ and Mn2 þ , but different amounts of distilled water to achieve the desired range of values for f. In addition to the matched series, 3H sources for each scintillant were prepared without any Fe3 þ or Mn2 þ in order to quantify quench effects owing to the presence of the metal ions. In both experiments, scintillant was dispensed via dispensette while distilled water was added by micropipette. In Experiment 1, carrier solution was added via pycnometer, while in Experiment 2, carrier solutions were added by micropipette. All other additions were performed gravimetrically via pycnometer. 20 mL high performance glass LS vials with poly-cone lined urea screw caps (PerkinElmer) were used in all experiments.
2.2. Dynamic light scattering DLS measurements were made using a Zetasizer Nano ZS (Malvern Instruments, Inc., Westborough, MA, USA), using the same protocols as in our previous work (Bergeron, 2012). Briefly, samples in quartz cuvettes were allowed to equilibrate to 20 1C for 60 s prior to three measurement cycles of ten 10 s scans (for a total measurement time of 300 s). Input values for viscosities and refractive indices (of the pure scintillants) were provided by PerkinElmer, and average hydrodynamic diameters (defined as the diameter of a hard sphere that diffuses at the same rate as the particle being measured) generated by the instrument software were recorded and uncertainties handled as described previously (Bergeron, 2012). All blank cocktails were measured.
CMn/(μg g 1)
f
0.02–0.17 0.02–0.17 0.02–0.17 0.09–0.17 0.10–0.17 0.09–0.17 0.50 0.54 0 or 0.50 0 or 0.50 0.50 0.54 0 or 0.50 0 or 0.50
0.03–0.16 0.05–0.17 0.05–0.17 0.05–0.17 0.04–0.18 0.05–0.17 0.05–0.17 0.05–0.17
2.3. Liquid scintillation counting efficiencies LS sources were counted on three commercial instruments: a Beckman LS6500, a Packard Tri-Carb A2500 TR (PerkinElmer, USA), and a Wallac 1414 Winspectral (PerkinElmer, USA) instrument. The different operating conditions of the three counters helps to eliminate instrumental idiosyncrasies as a source of bias in the comparison of counting efficiencies (Laureano-Pérez et al., 2007). After preparation, sources were dark adapted for at least 1 h before counting. After three counting cycles, the blanks were removed for DLS measurements (vide supra). In Experiment 1, the initial cycles revealed an overall standard deviation on the count rate from all blanks of o0.05 counts per second ( E5%), and no systematic relationship between the blank count rate and the aqueous fraction, so subsequent measurement cycles included only a subset (2 for each cocktail) of the original 15 blanks. In Experiment 2, the standard deviation on the count rate in the initial cycles was 40.2 counts per second (E13%), and so all blanks were counted in subsequent cycles so that the appropriate matched blank could provide the background subtraction data for its complementary 3H, 54Mn, and 55Fe cocktails. 2.4. MICELLE2 calculations Historically, difficulties in calculating LS counting efficiencies for electron capture nuclides, especially those emitting low-energy Auger electrons, arose from reliance on overly simple models for atomic rearrangement processes (Günther, 1998; Kossert and Grau Carles, 2006, 2010). Recently, the stochastic approach to atomic rearrangement as implemented in the MICELLE2 code has produced better agreement with experimental results, reducing typical discrepancies with experiment to r2% (Kossert and Grau Carles, 2010). In this study, MICELLE2 was run for each specific cocktail formulation. In Experiment 1, calculations were run using the two models for source distribution within the micelles (central and random) and ignoring the micelle effect. The actual difference between the corrections calculated by the central model and the random model is very small (r3 10-4 counts per decay in this study), and so in Experiment 2, the central distribution model was dropped. For each sample, simulations (each with 5 104 events) were run with the appropriate micelle radius and percent added water entered into the CTL.DAT and EFFCOMP.DAT files, respectively. The 3H efficiencies from the H3X.TAB output file are based on an analytical method and were not used in this study; instead, separate simulations were run for each specific 3H
284
D.E. Bergeron, L. Laureano-Pérez / Applied Radiation and Isotopes 87 (2014) 282–286
cocktail, applying the stochastic method in the same way as for the 55 Fe or 54Mn cocktails. A total of 63 simulations, 3 for each of the 21 cocktails containing either 3H or 55Fe, were run for Experiment 1. 78 simulations were run for Experiment 2, 2 for each of the 33 cocktails containing 3H, 55Fe, or 54Mn, plus 2 additional calculations for 6 cocktails to probe the sensitivity of the model to varying values for the Birks parameter, kB. A value of kB¼0.012 cm MeV-1 (which has become a sort of de facto canonical value at NIST) was used for most of the calculations, and the atomic composition (entered into the SCINTI. DAT input file) for each scintillant was taken from the manufacturer's specifications (PerkinElmer, 2007). The output of each calculation was used to relate the measured quench indicating parameter (QIP) from each tracer (3H or 54Mn) cocktail to a figure of merit. With carefully matched cocktails, the relationship between the 3H or 54Mn counting efficiency and the QIP can be translated via the figure of merit to a relationship between the QIP and a theoretical value for the 55Fe counting efficiency. In an ordinary CNET experiment, this theoretical 55 Fe efficiency is applied to the experimental count rates to determine the sample activity. In this study, the sample activity is known, and the value of interest is the difference between the calculated and measured counting efficiencies. This difference is measured in terms of efficiency residuals, defined as (εcalc εmeas)/εmeas, where εcalc and εmeas are the calculated and measured 55Fe efficiencies, respectively.
3. Results and discussion 3.1. Dynamic light scattering
Micelle Hydrodynamic Diameter /nm
Fig. 1 shows the results of the DLS measurements in terms of the dependence of the micelle hydrodynamic diameter on total aqueous fraction. The results are in general accord with our previous measurements (Bergeron, 2012). In the two ionic surfactant-containing cocktails, HiSafe II and Ultima Gold, micelle size increases steadily with f in the range between E 0.05 and E0.17. The HiSafe II data points in Fig. 1 at f E0.024 deviate from the monotonic trend. In our previous study (Bergeron, 2012), the HiSafe II data closely resembled the data for Ultima Gold, and no deviation from the monotonic trend was observed. Measurements were also made on a HiSafe II sample without added Fe3 þ carrier, and the deviation, though slightly smaller, was still present. Thus, the presence of metal ions does not explain the anomaly, and we
5
4
3
2
0
0.05
0.1
0.15
f Fig. 1. Results of dynamic light scattering (DLS) measurements on liquid scintillation cocktails with varying total aqueous fraction, f. The black diamonds and gray asterisks represent Ultima Gold cocktails with and without Fe3 þ and Mn2 þ added, respectively. The gray squares correspond to HiSafe II cocktails with Fe3 þ added. The open and closed triangles correspond to metal-containing HiSafe III cocktails from Experiments 1 and 2, respectively, and the black X’s correspond to HiSafe III with no added metals. The lines are intended only to guide the eye.
unfortunately do not have sufficient information to comment on its origin. The HiSafe III data shows a minimum in the micelle size at fE 0.09, resulting in the “U”-shaped curve, consistent with our previous study (Bergeron, 2012). The curve shapes and actual micelle hydrodynamic diameters seem to be unperturbed by the presence of metal ions in these experiments. 3.2. Liquid scintillation counting The HiSafe II cocktails were found to be unstable over time. All samples showed a rapid decline in counting efficiency over the first day, and the lower f samples showed a continued steady decrease in counting efficiency, falling more than 0.11 counts per decay over a 15 day period. The HiSafe III and Ultima Gold samples did not show any pronounced trending in counting efficiency over time. Overall quenching of the cocktails in terms of the instruments’ quench indicating parameters (QIPs) trended monotonically with f in both HiSafe III and Ultima Gold. Through measurements on 3H sources with and without Fe3 þ and Mn2 þ carrier solution, we determined that metal quenching is minimal, affecting QIPs by o2.5%, resulting in o 0.2% change in 3H εmeas. Since QIPs trend monotonically with f, they also trend monotonically with micelle size across the Ultima Gold series, but not across the HiSafe III series. This is due to the shape of the curves in Fig. 1 and shows that the micelle size effect is not on the same scale as overall quenching effects, as expected. Fig. 2 shows measured counting efficiencies for Ultima Gold and HiSafe III cocktails containing 3H, 54Mn, and 55Fe. Data taken on the Packard Tri-Carb A2500 TR during Experiment 2 is presented, but the trends are very consistent on all counters. Panels a and c reveal that the measured counting efficiency trends monotonically with f (and thus with QIP). Panel d reveals that the trend is not monotonic as a function of micelle size; this is to be expected since the magnitude of the micelle size effect is expected to be smaller than other quenching effects (Grau Carles, 2007; Kossert and Grau Carles, 2010). The tracing experiments were performed in an attempt to help isolate micelle effects. Fig. 3 presents 3H efficiency tracing data in terms of 55Fe efficiency residuals ((εcalc εmeas)/εmeas). Residuals calculated with and without the micelle correction are shown, and it is clear that the magnitude of the micelle correction is small relative to the overall disagreement between the measured and calculated efficiencies. The apparent correlation between micelle size and residuals arises because micelle size and efficiency are both correlated with aqueous fraction (vide supra). The MICELLE2 model underestimates the 55Fe counting efficiency, but the magnitude of the underestimation varies with tracer efficiency (Kossert and Grau Carles, 2010), and thus with aqueous fraction. We wish to emphasize that this trend is not caused by the micelle size effect, as we have attempted to make clear by plotting the data points with and without the micelle correction together in Fig. 3. The model dependence – in this case measured as the sensitivity of the recovered 55Fe activity to changes in sample quenching, and thus εmeas – appears to be more pronounced for Ultima Gold than HiSafe III, probably because a wider range of efficiencies is covered in this experimental scheme. For cocktail compositions (f E0.1) that would be typical in a CNET experiment at NIST, the efficiency is underestimated by 2% to 3% (Fig. 3). This range is in reasonable accord with previous reports (Kossert and Grau Carles, 2010). While not shown here, tracing with 54Mn improves the modeldependence somewhat, as expected based on the report of Günther (1998). This improvement was observed in our data as a reduction in the dependence on kB. We found that changing kB
Counting Efficiency / counts per decay
D.E. Bergeron, L. Laureano-Pérez / Applied Radiation and Isotopes 87 (2014) 282–286
0.450
0.450
0.425
0.425
0.400
0.400
0.375
0
0.05
0.1
0.15
0.2
0.375
0.425
0.425
0.375
0.375
0.325
0.325
0.275
0.275 0
0.05
0.1
0.15
0.2
f
2
3.5
2.5
3.7
3
3.9
4.1
285
3.5
4.3
4.5
4
4.7
Micelle Hydrodynamic Diameter / nm
Fig. 2. Measured counting efficiencies for Ultima Gold (a and b) and HiSafe III (c and d) cocktails containing 3H (open and closed circles), 54Mn (open and closed triangles), and 55Fe (open and closed diamonds). The same measured efficiencies are plotted against the total aqueous fraction, f, in panels a and c and against micelle size in panels b and d.
(εcalc- εmeas ) /εmeas / %
0
-2
-4
-6
-8
2
2.5
3
3.5
4
4.5
5
Micelle Hydrodynamic Diameter / nm Fig. 3. Residual plot for 55Fe efficiencies calculated using MICELLE2 vs. micelle size. The gray triangles and open circles correspond to Ultima Gold cocktails with and without the micelle correction, respectively. Black diamonds and open squares correspond to HiSafe III samples from Experiment 1 with and without the micelle correction, respectively. All micelle corrections shown use the “random distribution” model. Uncertainty (k ¼ 1) bars are calculated from the standard deviation on repeated measurements, combined with standard uncertainties on masses and activity. The largest uncertainty component is consistently the uncertainty on the massic activity of the master solution. Although there is obvious correlation between micelle size and traced efficiency, it arises because micelle size and efficiency both change as a function of aqueous fraction.
from 0.0075 cm MeV 1 to 0.015 cm MeV 1 resulted in a 3.0% change in the average 3H-traced efficiency residuals, but only a 0.92% change in the average 54Mn-traced efficiency residuals. The reduction in model dependence was also obvious in this experiment from the reduction in the range of efficiency residuals. In Fig. 3, the residuals for the Ultima Gold data range from 1.3% to 6.9% over the range of aqueous fractions covered in these experiments; the difference between the minimum and maximum is the residual range (Δresidual) of 5.6%. In the 54Mn tracing experiment, Δresidual was only 4.0%. While the model dependence was reduced using 54Mn as a tracer, we encountered a complication. In the course of our experiments, we discovered unaccounted impurities in the 54Mn solution (the data in Fig. 2 includes an impurity correction). Gamma spectroscopy revealed the likely presence of 152Eu, but the unexpectedly high LS counting efficiencies for the 54Mn solution also strongly suggested the presence of another impurity, probably 55Fe, which would not be detected by gamma spectroscopy. These long-lived impurities are common in 54Mn solutions
(Ratel and Michotte, 2003; Kovář et al., 1989). For this reason, sources more than a few years old should be reevaluated for radionuclidic impurities, including an assay for 55Fe, prior to use as a tracer. This experiment was designed to maximize and isolate the micelle size effect so that it might be measured. Our largest micelle correction, calculated with MICELLE2, was 0.30%, using the “random distribution” model for the largest experimentally achievable micelle size. Even with the best model currently available and the best tracer available, model-dependences can be expected to result in errors in calculated 55Fe efficiencies on the order of several percent, much larger than even the largest micelle size correction. Furthermore, other sources of uncertainty such as count repeatability, reproducibility, or counter-to-counter variations are often larger than 0.3%. Considering these factors, it seems that the micelle size effect will be negligible in efficiency tracing experiments involving standard commercial scintillants.
4. Conclusions The magnitude of the micelle size effect is small relative to typical sources of uncertainty in 55Fe liquid scintillation counting experiments. Furthermore, model-dependence introduces errors in calculated efficiencies that are more than an order of magnitude larger than the largest micelle size effect that should be expected in an experiment involving commercial scintillants. We were unable to detect any reductions in LS counting efficiency that might be attributable to the micelle size effect. We can, however, draw several important conclusions from this study. Micelle sizes are not significantly impacted by the presence of Fe3 þ or Mn2 þ ions at the concentrations considered in this work. Ultima Gold and HiSafe III yield stable cocktails with the compositions described herein. The best results for both cocktails were achieved with f¼ 0.05 to 0.15. HiSafe II cocktails were unstable over time. Despite the fact that metal ions can be efficient fluorescence quenchers (Hariharan et al., 1997; Hariharan and Mishra, 1998), we found that metal quenching in these cocktails is minimal, resulting in a o0.2% reduction in the counting efficiencies. Use of 54Mn as the tracer to determine 55Fe efficiencies reduces model dependence, but long-lived impurities in an older 54 Mn source can complicate measurements. In our previous report (Bergeron, 2012), we showed that the micelle size effect should be trivially small in commercial LS cocktail formulations for all but the most sensitive low-energy Auger-emitting nuclides, such as 55Fe. In this work, we have
286
D.E. Bergeron, L. Laureano-Pérez / Applied Radiation and Isotopes 87 (2014) 282–286
demonstrated that even for 55Fe, the effect is too small to measure, and is dwarfed by model dependences and uncertainties. The best model currently available for calculating scintillation efficiencies for electron capture nuclides is MICELLE2 (Kossert and Grau Carles, 2010), and it allows for easy inclusion of the micelle correction. We hope that the micelle size data reported here and in our previous work (Bergeron, 2012) will be of use to researchers who wish to make the small corrections. While the micelle size effect may not substantially affect counting efficiencies in most experiments, there are other “micelle effects” that may yet prove important. The presence or absence of micelles in a LS cocktail can be expected to affect the transmission of optical photons, so that effects around the critical micelle concentration become very relevant to LS counting experiments. In general, accounting for scattering of optical photons by micelles may prove important (Santiago et al., 2013), especially in cases where the number of scintillation photons is low. In addition, chemical quenching may be influenced by the unique chemistry occurring at micellar interfaces. For example, Blach et al., 2011 used a solvatochromic probe technique to establish that changes to the hydrogen bond donor ability wrought by confinement interactions at the water-surfactant interface of reverse micelles can alter the electron donor properties of the water oxygen nonbonding electron pairs. We hope that future investigations will shed light on how these and other phenomena affect fluorescence quenching in LS cocktails. Finally, energy loss mechanisms analogous to the micelle size effect remain extremely important in the description of gel (Grau Carles, 2006, 2007) and plastic microsphere scintillation media (Tarancón Sanz and Kossert, 2011; Santiago et al., 2013). As attempts to model those systems progress, a full understanding of the energy-loss and light scattering mechanisms will become increasingly important. Acknowledgements We are grateful to Jeffrey T. Cessna for illuminating discussions, to Ron Collé for an insightful critique of the manuscript, to Rebecca Zangmeister for access to the DLS instrument, and to Augustín Grau Carles for provision of the MICELLE2 code. References Andrade, S.M., Costa, M.B., 2002. The aqueous environment in AOT and Triton X100 (w/o) microemulsions probed by fluorescence. Photochem. Photobiol. Sci. 1, 500–506. Bergeron, D.E., 2012. Determination of micelle size in some commercial liquid scintillation cocktails. Appl. Radiat. Isot 70, 2164–2169.
Blach, D., Mariano Correa, N., Silber, J.J., Darío Falcone, R., 2011. Interfacial water with special electron donor properties: effect of water-surfactant interaction in confined reversed micellar environments and its influence on the coordination chemistry of a copper complex. J. Colloid Interface Sci 355, 124–130. Faeder, J., Ladanyi, B.M., 2000. Molecular dynamics simulations of the interior of aqueous reverse micelles. J. Phys. Chem. B 104, 1033–1046. Faeder, J., Ladanyi, B.M., 2001. Solvation dynamics in aqueous reverse micelles: a computer simulation study. J. Phys. Chem. B 105, 11148–11158. Faeder, J., Ladanyi, B.M., 2005. Solvation dynamics in reverse micelles: the role of headgroup-solute interactions. J. Phys. Chem. B 109, 6732–6740. Grau Carles, A., 2006. Nanodosemeters based on gel scintillators. Radiat. Prot. Dosim. 122, 420–426. Grau Carles, A., 2007. MICELLE, the micelle size effect on the LS counting efficiency. Comput. Phys. Commun. 176, 305–317. Günther, E., 1998. Standardization of the EC nuclides 55Fe and 65Zn with the CIEMAT/NIST LSC tracer method. Appl. Radiat. Isot 49, 1055–1060. Hariharan, C., Vijaysree, V., Mishra, A.K., 1997. Quenching of 2,5-diphenyloxazole (PPO) fluorescence by metal ions. J. Lumin. 75, 205–211. Hariharan, C., Mishra, A.K., 1998. Fluorescence quenching of liquid scintillation counters by metal ions. Radiat. Meas. 29, 473–480. Kossert, K., Grau Carles, A., 2006. The LSC efficiency for low-Z electron-capture nuclides. Appl. Radiat. Isot 64, 1446–1453. Kossert, K., Grau Carles, A., 2008. Study of a Monte Carlo rearrangement model for the activity determination of electron-capture nuclides by means of liquid scintillation counting. Appl. Radiat. Isot 66, 998–1005. Kossert, K., Grau Carles, A., 2010. Improved method for the calculation of the counting efficiency of electron-capture nuclides in liquid scintillation samples. Appl. Radiat. Isot 68, 1482–1488. Kovář, P., Dryák, P., Plchová, L., Kokta, L., Novotná, P., 1989. Impurities in Radionuclide Materials II. Revised Edition. Institute of Research, Production and Application of Radioisotopes, Prague. Laureano-Pérez, L., Collé, R., Fitzgerald, R., Outola, I., Pibida, L., 2007. A liquidscintillation-based primary standardization of 210Pb. Appl. Radiat. Isot 65, 1368–1380. NIST, 2006. Standard Reference Material 4929F: Iron-55 Radioactivity Standard. NIST, 2008. Standard Reference Material 4927F: Hydrogen-3 Radioactivity Standard. Pant, D., Riter, R.E., Levinger, N.E., 1998. Influence of restricted environment and ionic interactions on water solvation dynamics. J. Chem. Phys. 109, 9995–10003. PerkinElmer, 2007. LSC in Practice: LSC Cocktails—Elemental Composition, Application Note, 〈http://www.perkinelmer.com/CMSResources/Images/44-73093APP_LSCCock tailsElementalComposition.pdf〉. Pileni, M.-P., Zemb, T., Petit, C., 1985. Solubilization by reverse micelles: solute localization and structure perturbation. Chem. Phys. Lett. 118, 414–420. Ratel, G., Michotte, C., 2003. BIPM Comparison BIPM.RI(II)-K1.Mn-54 of Activity Measurements of the Radionuclide 54Mn . Metrologia 40, Tech. Supp., 06025, http://dx.doi.org/10.1088/0026-1394/40/1A/06025. Rodríguez, R., Vargas, S., Fernández-Velasco, D.A., 1998. Reverse micelle systems composed of water, Triton X-100, and phospholipids in organic solvents—1. Phase boundary titrations and dynamic light scattering analysis. J. Colloid. Interface Sci 197, 21–28. Santiago, L.M., Bagán, H., Tarancón, A., Rauret, G., Garcia, J.F., 2013. Systematic study of particle quenching in organic scintillators. Nucl. Instrum. Methods Phys. Res., Sect. A 698, 26–36. Stahla, M.L., Baruah, B., James, D.M., Johnson, M.D., Levinger, N.E., Crans, D.C., 2008. 1 H NMR studies of aerosol-OT reverse micelles with alkali and magnesium counterions: preparation and analysis of MAOTs. Langmuir 24, 6027–6035. Tarancón Sanz, A., Kossert, K., 2011. Application of a free parameter model to plastic scintillation samples. Nucl. Instrum. Methods Phys. Res., Sect. A 648, 124–131.
Contents lists available at ScienceDirect
Applied Radiation and Isotopes journal homepage: www.elsevier.com/locate/apradiso
Micelle size effect on Fe-55 liquid scintillation efficiency Denis E. Bergeron n, Lizbeth Laureano-Pérez Physical Measurements Laboratory, National Institute of Standards and Technology, 100 Bureau Dr., Gaithersburg, MD 20899-8462, USA
H I G H L I G H T S
No efficiency reductions attributable to the micelle size effect were observed. Calculated micelle corrections are small compared to other uncertainties. Metal ions did not affect micelle size or fluorescence quenching. Tracing 55Fe efficiencies with 54Mn reduced model dependence.
art ic l e i nf o
a b s t r a c t
Available online 1 December 2013
We used efficiency tracing techniques to study the micelle size effect on liquid scintillation counting of the electron capture nuclide, 55Fe. We determined micelle hydrodynamic diameters for specific LS cocktails via dynamic light scattering, and sought trends in efficiencies as a function of micelle size. The presence of Fe3 þ or Mn2 þ ions in the cocktails did not significantly affect micelle sizes or fluorescence quenching. We did not detect any reductions in counting efficiencies due to the micelle size effect. Published by Elsevier Ltd.
Keywords: Reverse micelle Liquid scintillation counting Cocktail Mn-54 H-3
1. Introduction We recently undertook a series of dynamic light scattering (DLS) measurements to determine the size of micelles in several commercial liquid scintillation (LS) cocktails (Bergeron, 2012). An electron emitted from a radionuclide residing within a reverse micelle will lose energy to the aqueous medium before interacting with the scintillant residing in the organic phase, resulting in a reduction in counting efficiency typically referred to as “the micelle effect” or “micelle size effect” (Grau Carles, 2007; Kossert and Grau Carles, 2008, 2010). Our measurements revealed that micelles in commercial LS cocktails are smaller than previously assumed (based on measurements by Rodríguez et al., 1998), so that the micelle size effect is smaller than was thought. We showed that even for the electron capture nuclide 55Fe, the effect would reduce counting efficiencies by o0.2% under typical counting conditions (Bergeron, 2012). In this work, we set out to determine whether it might be possible to observe the micelle size effect under ordinary experimental conditions. We selected 55Fe because, as has been described previously (Grau Carles, 2006, 2007; Kossert and Grau Carles, 2010; Bergeron, 2012), its low energy Auger electron emissions make it particularly sensitive to the effect. Through careful cocktail selection, we sought to maximize the achievable range of micelle sizes. We were also careful to select cocktails with ionic and nonionic surfactants since the
n
Corresponding author. Tel.: þ 1 301 975 2282; fax: þ1 301 926 7416. E-mail addresses: [email protected], [email protected] (D.E. Bergeron).
0969-8043/$ - see front matter Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.apradiso.2013.11.080
location of the nuclide within the micelle, which may be affected by the surfactant type (Pileni et al., 1985; Pant et al., 1998; Andrade and Costa, 2002; Faeder and Ladanyi, 2000, 2001, 2005; Stahla et al., 2008), should affect the magnitude of the efficiency reduction. Since the variation of micelle size is achieved by varying the total aqueous fraction in the cocktail, and since the addition of water changes the general quenching properties of the cocktail, the experimental scheme included CIEMAT/NIST efficiency tracing (CNET) with both 3H and 54 Mn. It was hoped that careful efficiency tracing would help to minimize contributions from quenching mechanisms that might obscure the detection of the micelle size effect.
2. Methods 2.1. Sample preparation Two experiments were performed. All of the selected scintillants were diisopropyl naphthalene (DIN) based. Table 1 provides details on the scintillation cocktails used in both experiments. In Experiment 1, cocktails were prepared from HiSafe II and HiSafe III (PerkinElmer, Waltham, MA, USA).1 The scintillants were 1 Certain commercial equipment, instruments, or materials are identified in this paper to foster understanding. Such identification does not imply recommendation by the National Institute of Standards and Technology, nor does it imply that the materials or equipment identified are necessarily the best available for the purpose.
D.E. Bergeron, L. Laureano-Pérez / Applied Radiation and Isotopes 87 (2014) 282–286
283
Table 1 Summary of the LS cocktails used in the experiments. Each series includes NS separate samples, each with a different total aqueous fraction, f, covering the indicated range. CFe and CMn indicate the total concentration of the metal ions; the metal ions are introduced as carrier solution, or (indicated by italicization) standard solutions of the appropriate radionuclide. For more details on the surfactants and other components of the scintillants, refer to the manufacturer's specifications and material safety data sheets. Radionuclide solution Experiment 1 Fe H Blank 55 Fe 3 H Blank Experiment 2 55 Fe 54 Mn 3 H Blank 55 Fe 54 Mn 3 H Blank 55
Scintillant
Surfactant information
Ns
CFe/(μg g 1)
HiSafe II
In addition to nonionic surfactants, includes 10–20% sodium dioctyl sulphosuccinate (anionic surfactant)
HiSafe III
20% to 40% ethoxylate polymers (nonionic surfactants)
6 6 8 5 5 7
0.065 0.065 0 or 0.065 0.065 0.065 0 or 0.065
Ultima Gold
In addition to nonionic surfactants, indludes r 2.5% sodium dioctyl sulphosuccinate (anionic surfactant)
HiSafe III
20–40% Ethoxylate polymers (nonionic surfactants)
6 6 8 7 5 5 7 6
0.67 0.67 0 or 0.67 0 or 0.67 0.67 0.67 0 or 0.67 0 or 0.67
3
selected to include one formulation that contains an ionic surfactant, and one that includes only non-ionic surfactants (Table 1). Matched 55Fe, 3H, and blank cocktails were prepared with a range of aqueous fractions (f¼ vw/vtot; where vw is the total aqueous sample volume and vtot is the total cocktail volume, including added aqueous sample) selected to maximize the range of micelle sizes as measured by DLS in terms of hydrodynamic diameter (Bergeron, 2012). The 55Fe cocktails were prepared gravimetrically from a 14.2 kBq g-1 55Fe solution with 19 μg g 1 of Fe3 þ (as FeCl3) in 1 mol L 1 HCl (NIST, 2006). The matched 3H (NIST, 2008) cocktails and blanks contained Fe3 þ carrier solution. While the Fe3 þ content for all cocktails was the same, the desired range of f was achieved via addition of different amounts of distilled water, so that the cocktails with highest f had the lowest Fe3 þ concentration. In Experiment 2, HiSafe II was replaced by Ultima Gold (PerkinElmer, Waltham, MA, USA). All cocktails contained the same total amount of Fe3 þ and Mn2 þ , but different amounts of distilled water to achieve the desired range of values for f. In addition to the matched series, 3H sources for each scintillant were prepared without any Fe3 þ or Mn2 þ in order to quantify quench effects owing to the presence of the metal ions. In both experiments, scintillant was dispensed via dispensette while distilled water was added by micropipette. In Experiment 1, carrier solution was added via pycnometer, while in Experiment 2, carrier solutions were added by micropipette. All other additions were performed gravimetrically via pycnometer. 20 mL high performance glass LS vials with poly-cone lined urea screw caps (PerkinElmer) were used in all experiments.
2.2. Dynamic light scattering DLS measurements were made using a Zetasizer Nano ZS (Malvern Instruments, Inc., Westborough, MA, USA), using the same protocols as in our previous work (Bergeron, 2012). Briefly, samples in quartz cuvettes were allowed to equilibrate to 20 1C for 60 s prior to three measurement cycles of ten 10 s scans (for a total measurement time of 300 s). Input values for viscosities and refractive indices (of the pure scintillants) were provided by PerkinElmer, and average hydrodynamic diameters (defined as the diameter of a hard sphere that diffuses at the same rate as the particle being measured) generated by the instrument software were recorded and uncertainties handled as described previously (Bergeron, 2012). All blank cocktails were measured.
CMn/(μg g 1)
f
0.02–0.17 0.02–0.17 0.02–0.17 0.09–0.17 0.10–0.17 0.09–0.17 0.50 0.54 0 or 0.50 0 or 0.50 0.50 0.54 0 or 0.50 0 or 0.50
0.03–0.16 0.05–0.17 0.05–0.17 0.05–0.17 0.04–0.18 0.05–0.17 0.05–0.17 0.05–0.17
2.3. Liquid scintillation counting efficiencies LS sources were counted on three commercial instruments: a Beckman LS6500, a Packard Tri-Carb A2500 TR (PerkinElmer, USA), and a Wallac 1414 Winspectral (PerkinElmer, USA) instrument. The different operating conditions of the three counters helps to eliminate instrumental idiosyncrasies as a source of bias in the comparison of counting efficiencies (Laureano-Pérez et al., 2007). After preparation, sources were dark adapted for at least 1 h before counting. After three counting cycles, the blanks were removed for DLS measurements (vide supra). In Experiment 1, the initial cycles revealed an overall standard deviation on the count rate from all blanks of o0.05 counts per second ( E5%), and no systematic relationship between the blank count rate and the aqueous fraction, so subsequent measurement cycles included only a subset (2 for each cocktail) of the original 15 blanks. In Experiment 2, the standard deviation on the count rate in the initial cycles was 40.2 counts per second (E13%), and so all blanks were counted in subsequent cycles so that the appropriate matched blank could provide the background subtraction data for its complementary 3H, 54Mn, and 55Fe cocktails. 2.4. MICELLE2 calculations Historically, difficulties in calculating LS counting efficiencies for electron capture nuclides, especially those emitting low-energy Auger electrons, arose from reliance on overly simple models for atomic rearrangement processes (Günther, 1998; Kossert and Grau Carles, 2006, 2010). Recently, the stochastic approach to atomic rearrangement as implemented in the MICELLE2 code has produced better agreement with experimental results, reducing typical discrepancies with experiment to r2% (Kossert and Grau Carles, 2010). In this study, MICELLE2 was run for each specific cocktail formulation. In Experiment 1, calculations were run using the two models for source distribution within the micelles (central and random) and ignoring the micelle effect. The actual difference between the corrections calculated by the central model and the random model is very small (r3 10-4 counts per decay in this study), and so in Experiment 2, the central distribution model was dropped. For each sample, simulations (each with 5 104 events) were run with the appropriate micelle radius and percent added water entered into the CTL.DAT and EFFCOMP.DAT files, respectively. The 3H efficiencies from the H3X.TAB output file are based on an analytical method and were not used in this study; instead, separate simulations were run for each specific 3H
284
D.E. Bergeron, L. Laureano-Pérez / Applied Radiation and Isotopes 87 (2014) 282–286
cocktail, applying the stochastic method in the same way as for the 55 Fe or 54Mn cocktails. A total of 63 simulations, 3 for each of the 21 cocktails containing either 3H or 55Fe, were run for Experiment 1. 78 simulations were run for Experiment 2, 2 for each of the 33 cocktails containing 3H, 55Fe, or 54Mn, plus 2 additional calculations for 6 cocktails to probe the sensitivity of the model to varying values for the Birks parameter, kB. A value of kB¼0.012 cm MeV-1 (which has become a sort of de facto canonical value at NIST) was used for most of the calculations, and the atomic composition (entered into the SCINTI. DAT input file) for each scintillant was taken from the manufacturer's specifications (PerkinElmer, 2007). The output of each calculation was used to relate the measured quench indicating parameter (QIP) from each tracer (3H or 54Mn) cocktail to a figure of merit. With carefully matched cocktails, the relationship between the 3H or 54Mn counting efficiency and the QIP can be translated via the figure of merit to a relationship between the QIP and a theoretical value for the 55Fe counting efficiency. In an ordinary CNET experiment, this theoretical 55 Fe efficiency is applied to the experimental count rates to determine the sample activity. In this study, the sample activity is known, and the value of interest is the difference between the calculated and measured counting efficiencies. This difference is measured in terms of efficiency residuals, defined as (εcalc εmeas)/εmeas, where εcalc and εmeas are the calculated and measured 55Fe efficiencies, respectively.
3. Results and discussion 3.1. Dynamic light scattering
Micelle Hydrodynamic Diameter /nm
Fig. 1 shows the results of the DLS measurements in terms of the dependence of the micelle hydrodynamic diameter on total aqueous fraction. The results are in general accord with our previous measurements (Bergeron, 2012). In the two ionic surfactant-containing cocktails, HiSafe II and Ultima Gold, micelle size increases steadily with f in the range between E 0.05 and E0.17. The HiSafe II data points in Fig. 1 at f E0.024 deviate from the monotonic trend. In our previous study (Bergeron, 2012), the HiSafe II data closely resembled the data for Ultima Gold, and no deviation from the monotonic trend was observed. Measurements were also made on a HiSafe II sample without added Fe3 þ carrier, and the deviation, though slightly smaller, was still present. Thus, the presence of metal ions does not explain the anomaly, and we
5
4
3
2
0
0.05
0.1
0.15
f Fig. 1. Results of dynamic light scattering (DLS) measurements on liquid scintillation cocktails with varying total aqueous fraction, f. The black diamonds and gray asterisks represent Ultima Gold cocktails with and without Fe3 þ and Mn2 þ added, respectively. The gray squares correspond to HiSafe II cocktails with Fe3 þ added. The open and closed triangles correspond to metal-containing HiSafe III cocktails from Experiments 1 and 2, respectively, and the black X’s correspond to HiSafe III with no added metals. The lines are intended only to guide the eye.
unfortunately do not have sufficient information to comment on its origin. The HiSafe III data shows a minimum in the micelle size at fE 0.09, resulting in the “U”-shaped curve, consistent with our previous study (Bergeron, 2012). The curve shapes and actual micelle hydrodynamic diameters seem to be unperturbed by the presence of metal ions in these experiments. 3.2. Liquid scintillation counting The HiSafe II cocktails were found to be unstable over time. All samples showed a rapid decline in counting efficiency over the first day, and the lower f samples showed a continued steady decrease in counting efficiency, falling more than 0.11 counts per decay over a 15 day period. The HiSafe III and Ultima Gold samples did not show any pronounced trending in counting efficiency over time. Overall quenching of the cocktails in terms of the instruments’ quench indicating parameters (QIPs) trended monotonically with f in both HiSafe III and Ultima Gold. Through measurements on 3H sources with and without Fe3 þ and Mn2 þ carrier solution, we determined that metal quenching is minimal, affecting QIPs by o2.5%, resulting in o 0.2% change in 3H εmeas. Since QIPs trend monotonically with f, they also trend monotonically with micelle size across the Ultima Gold series, but not across the HiSafe III series. This is due to the shape of the curves in Fig. 1 and shows that the micelle size effect is not on the same scale as overall quenching effects, as expected. Fig. 2 shows measured counting efficiencies for Ultima Gold and HiSafe III cocktails containing 3H, 54Mn, and 55Fe. Data taken on the Packard Tri-Carb A2500 TR during Experiment 2 is presented, but the trends are very consistent on all counters. Panels a and c reveal that the measured counting efficiency trends monotonically with f (and thus with QIP). Panel d reveals that the trend is not monotonic as a function of micelle size; this is to be expected since the magnitude of the micelle size effect is expected to be smaller than other quenching effects (Grau Carles, 2007; Kossert and Grau Carles, 2010). The tracing experiments were performed in an attempt to help isolate micelle effects. Fig. 3 presents 3H efficiency tracing data in terms of 55Fe efficiency residuals ((εcalc εmeas)/εmeas). Residuals calculated with and without the micelle correction are shown, and it is clear that the magnitude of the micelle correction is small relative to the overall disagreement between the measured and calculated efficiencies. The apparent correlation between micelle size and residuals arises because micelle size and efficiency are both correlated with aqueous fraction (vide supra). The MICELLE2 model underestimates the 55Fe counting efficiency, but the magnitude of the underestimation varies with tracer efficiency (Kossert and Grau Carles, 2010), and thus with aqueous fraction. We wish to emphasize that this trend is not caused by the micelle size effect, as we have attempted to make clear by plotting the data points with and without the micelle correction together in Fig. 3. The model dependence – in this case measured as the sensitivity of the recovered 55Fe activity to changes in sample quenching, and thus εmeas – appears to be more pronounced for Ultima Gold than HiSafe III, probably because a wider range of efficiencies is covered in this experimental scheme. For cocktail compositions (f E0.1) that would be typical in a CNET experiment at NIST, the efficiency is underestimated by 2% to 3% (Fig. 3). This range is in reasonable accord with previous reports (Kossert and Grau Carles, 2010). While not shown here, tracing with 54Mn improves the modeldependence somewhat, as expected based on the report of Günther (1998). This improvement was observed in our data as a reduction in the dependence on kB. We found that changing kB
Counting Efficiency / counts per decay
D.E. Bergeron, L. Laureano-Pérez / Applied Radiation and Isotopes 87 (2014) 282–286
0.450
0.450
0.425
0.425
0.400
0.400
0.375
0
0.05
0.1
0.15
0.2
0.375
0.425
0.425
0.375
0.375
0.325
0.325
0.275
0.275 0
0.05
0.1
0.15
0.2
f
2
3.5
2.5
3.7
3
3.9
4.1
285
3.5
4.3
4.5
4
4.7
Micelle Hydrodynamic Diameter / nm
Fig. 2. Measured counting efficiencies for Ultima Gold (a and b) and HiSafe III (c and d) cocktails containing 3H (open and closed circles), 54Mn (open and closed triangles), and 55Fe (open and closed diamonds). The same measured efficiencies are plotted against the total aqueous fraction, f, in panels a and c and against micelle size in panels b and d.
(εcalc- εmeas ) /εmeas / %
0
-2
-4
-6
-8
2
2.5
3
3.5
4
4.5
5
Micelle Hydrodynamic Diameter / nm Fig. 3. Residual plot for 55Fe efficiencies calculated using MICELLE2 vs. micelle size. The gray triangles and open circles correspond to Ultima Gold cocktails with and without the micelle correction, respectively. Black diamonds and open squares correspond to HiSafe III samples from Experiment 1 with and without the micelle correction, respectively. All micelle corrections shown use the “random distribution” model. Uncertainty (k ¼ 1) bars are calculated from the standard deviation on repeated measurements, combined with standard uncertainties on masses and activity. The largest uncertainty component is consistently the uncertainty on the massic activity of the master solution. Although there is obvious correlation between micelle size and traced efficiency, it arises because micelle size and efficiency both change as a function of aqueous fraction.
from 0.0075 cm MeV 1 to 0.015 cm MeV 1 resulted in a 3.0% change in the average 3H-traced efficiency residuals, but only a 0.92% change in the average 54Mn-traced efficiency residuals. The reduction in model dependence was also obvious in this experiment from the reduction in the range of efficiency residuals. In Fig. 3, the residuals for the Ultima Gold data range from 1.3% to 6.9% over the range of aqueous fractions covered in these experiments; the difference between the minimum and maximum is the residual range (Δresidual) of 5.6%. In the 54Mn tracing experiment, Δresidual was only 4.0%. While the model dependence was reduced using 54Mn as a tracer, we encountered a complication. In the course of our experiments, we discovered unaccounted impurities in the 54Mn solution (the data in Fig. 2 includes an impurity correction). Gamma spectroscopy revealed the likely presence of 152Eu, but the unexpectedly high LS counting efficiencies for the 54Mn solution also strongly suggested the presence of another impurity, probably 55Fe, which would not be detected by gamma spectroscopy. These long-lived impurities are common in 54Mn solutions
(Ratel and Michotte, 2003; Kovář et al., 1989). For this reason, sources more than a few years old should be reevaluated for radionuclidic impurities, including an assay for 55Fe, prior to use as a tracer. This experiment was designed to maximize and isolate the micelle size effect so that it might be measured. Our largest micelle correction, calculated with MICELLE2, was 0.30%, using the “random distribution” model for the largest experimentally achievable micelle size. Even with the best model currently available and the best tracer available, model-dependences can be expected to result in errors in calculated 55Fe efficiencies on the order of several percent, much larger than even the largest micelle size correction. Furthermore, other sources of uncertainty such as count repeatability, reproducibility, or counter-to-counter variations are often larger than 0.3%. Considering these factors, it seems that the micelle size effect will be negligible in efficiency tracing experiments involving standard commercial scintillants.
4. Conclusions The magnitude of the micelle size effect is small relative to typical sources of uncertainty in 55Fe liquid scintillation counting experiments. Furthermore, model-dependence introduces errors in calculated efficiencies that are more than an order of magnitude larger than the largest micelle size effect that should be expected in an experiment involving commercial scintillants. We were unable to detect any reductions in LS counting efficiency that might be attributable to the micelle size effect. We can, however, draw several important conclusions from this study. Micelle sizes are not significantly impacted by the presence of Fe3 þ or Mn2 þ ions at the concentrations considered in this work. Ultima Gold and HiSafe III yield stable cocktails with the compositions described herein. The best results for both cocktails were achieved with f¼ 0.05 to 0.15. HiSafe II cocktails were unstable over time. Despite the fact that metal ions can be efficient fluorescence quenchers (Hariharan et al., 1997; Hariharan and Mishra, 1998), we found that metal quenching in these cocktails is minimal, resulting in a o0.2% reduction in the counting efficiencies. Use of 54Mn as the tracer to determine 55Fe efficiencies reduces model dependence, but long-lived impurities in an older 54 Mn source can complicate measurements. In our previous report (Bergeron, 2012), we showed that the micelle size effect should be trivially small in commercial LS cocktail formulations for all but the most sensitive low-energy Auger-emitting nuclides, such as 55Fe. In this work, we have
286
D.E. Bergeron, L. Laureano-Pérez / Applied Radiation and Isotopes 87 (2014) 282–286
demonstrated that even for 55Fe, the effect is too small to measure, and is dwarfed by model dependences and uncertainties. The best model currently available for calculating scintillation efficiencies for electron capture nuclides is MICELLE2 (Kossert and Grau Carles, 2010), and it allows for easy inclusion of the micelle correction. We hope that the micelle size data reported here and in our previous work (Bergeron, 2012) will be of use to researchers who wish to make the small corrections. While the micelle size effect may not substantially affect counting efficiencies in most experiments, there are other “micelle effects” that may yet prove important. The presence or absence of micelles in a LS cocktail can be expected to affect the transmission of optical photons, so that effects around the critical micelle concentration become very relevant to LS counting experiments. In general, accounting for scattering of optical photons by micelles may prove important (Santiago et al., 2013), especially in cases where the number of scintillation photons is low. In addition, chemical quenching may be influenced by the unique chemistry occurring at micellar interfaces. For example, Blach et al., 2011 used a solvatochromic probe technique to establish that changes to the hydrogen bond donor ability wrought by confinement interactions at the water-surfactant interface of reverse micelles can alter the electron donor properties of the water oxygen nonbonding electron pairs. We hope that future investigations will shed light on how these and other phenomena affect fluorescence quenching in LS cocktails. Finally, energy loss mechanisms analogous to the micelle size effect remain extremely important in the description of gel (Grau Carles, 2006, 2007) and plastic microsphere scintillation media (Tarancón Sanz and Kossert, 2011; Santiago et al., 2013). As attempts to model those systems progress, a full understanding of the energy-loss and light scattering mechanisms will become increasingly important. Acknowledgements We are grateful to Jeffrey T. Cessna for illuminating discussions, to Ron Collé for an insightful critique of the manuscript, to Rebecca Zangmeister for access to the DLS instrument, and to Augustín Grau Carles for provision of the MICELLE2 code. References Andrade, S.M., Costa, M.B., 2002. The aqueous environment in AOT and Triton X100 (w/o) microemulsions probed by fluorescence. Photochem. Photobiol. Sci. 1, 500–506. Bergeron, D.E., 2012. Determination of micelle size in some commercial liquid scintillation cocktails. Appl. Radiat. Isot 70, 2164–2169.
Blach, D., Mariano Correa, N., Silber, J.J., Darío Falcone, R., 2011. Interfacial water with special electron donor properties: effect of water-surfactant interaction in confined reversed micellar environments and its influence on the coordination chemistry of a copper complex. J. Colloid Interface Sci 355, 124–130. Faeder, J., Ladanyi, B.M., 2000. Molecular dynamics simulations of the interior of aqueous reverse micelles. J. Phys. Chem. B 104, 1033–1046. Faeder, J., Ladanyi, B.M., 2001. Solvation dynamics in aqueous reverse micelles: a computer simulation study. J. Phys. Chem. B 105, 11148–11158. Faeder, J., Ladanyi, B.M., 2005. Solvation dynamics in reverse micelles: the role of headgroup-solute interactions. J. Phys. Chem. B 109, 6732–6740. Grau Carles, A., 2006. Nanodosemeters based on gel scintillators. Radiat. Prot. Dosim. 122, 420–426. Grau Carles, A., 2007. MICELLE, the micelle size effect on the LS counting efficiency. Comput. Phys. Commun. 176, 305–317. Günther, E., 1998. Standardization of the EC nuclides 55Fe and 65Zn with the CIEMAT/NIST LSC tracer method. Appl. Radiat. Isot 49, 1055–1060. Hariharan, C., Vijaysree, V., Mishra, A.K., 1997. Quenching of 2,5-diphenyloxazole (PPO) fluorescence by metal ions. J. Lumin. 75, 205–211. Hariharan, C., Mishra, A.K., 1998. Fluorescence quenching of liquid scintillation counters by metal ions. Radiat. Meas. 29, 473–480. Kossert, K., Grau Carles, A., 2006. The LSC efficiency for low-Z electron-capture nuclides. Appl. Radiat. Isot 64, 1446–1453. Kossert, K., Grau Carles, A., 2008. Study of a Monte Carlo rearrangement model for the activity determination of electron-capture nuclides by means of liquid scintillation counting. Appl. Radiat. Isot 66, 998–1005. Kossert, K., Grau Carles, A., 2010. Improved method for the calculation of the counting efficiency of electron-capture nuclides in liquid scintillation samples. Appl. Radiat. Isot 68, 1482–1488. Kovář, P., Dryák, P., Plchová, L., Kokta, L., Novotná, P., 1989. Impurities in Radionuclide Materials II. Revised Edition. Institute of Research, Production and Application of Radioisotopes, Prague. Laureano-Pérez, L., Collé, R., Fitzgerald, R., Outola, I., Pibida, L., 2007. A liquidscintillation-based primary standardization of 210Pb. Appl. Radiat. Isot 65, 1368–1380. NIST, 2006. Standard Reference Material 4929F: Iron-55 Radioactivity Standard. NIST, 2008. Standard Reference Material 4927F: Hydrogen-3 Radioactivity Standard. Pant, D., Riter, R.E., Levinger, N.E., 1998. Influence of restricted environment and ionic interactions on water solvation dynamics. J. Chem. Phys. 109, 9995–10003. PerkinElmer, 2007. LSC in Practice: LSC Cocktails—Elemental Composition, Application Note, 〈http://www.perkinelmer.com/CMSResources/Images/44-73093APP_LSCCock tailsElementalComposition.pdf〉. Pileni, M.-P., Zemb, T., Petit, C., 1985. Solubilization by reverse micelles: solute localization and structure perturbation. Chem. Phys. Lett. 118, 414–420. Ratel, G., Michotte, C., 2003. BIPM Comparison BIPM.RI(II)-K1.Mn-54 of Activity Measurements of the Radionuclide 54Mn . Metrologia 40, Tech. Supp., 06025, http://dx.doi.org/10.1088/0026-1394/40/1A/06025. Rodríguez, R., Vargas, S., Fernández-Velasco, D.A., 1998. Reverse micelle systems composed of water, Triton X-100, and phospholipids in organic solvents—1. Phase boundary titrations and dynamic light scattering analysis. J. Colloid. Interface Sci 197, 21–28. Santiago, L.M., Bagán, H., Tarancón, A., Rauret, G., Garcia, J.F., 2013. Systematic study of particle quenching in organic scintillators. Nucl. Instrum. Methods Phys. Res., Sect. A 698, 26–36. Stahla, M.L., Baruah, B., James, D.M., Johnson, M.D., Levinger, N.E., Crans, D.C., 2008. 1 H NMR studies of aerosol-OT reverse micelles with alkali and magnesium counterions: preparation and analysis of MAOTs. Langmuir 24, 6027–6035. Tarancón Sanz, A., Kossert, K., 2011. Application of a free parameter model to plastic scintillation samples. Nucl. Instrum. Methods Phys. Res., Sect. A 648, 124–131.
