Answer to Question #3760 Submitted to "Ask the Experts"

My questions are about measuring ¹⁰⁹Cd and ¹⁴C in the same sample.

1. Is there any way that I can measure both radionuclides simultaneously on a scintillation counter?

2. If not, do you think it is possible to first count the sample in a gamma counter (which should only pick up the ¹⁰⁹Cd), then from that work out the amount of Cd present, then run the same sample in the scintillation counter and subtract the amount of Cd measured in the gamma counter from the total radioactivity detected by the scintillation counter? For your information, I have run some standards and see that the ¹⁰⁹Cd and ¹⁴C spectra overlap almost exactly: the ¹⁰⁹Cd has a double-peaked spectrum—one sharper peak at about 5 keV, and then a broader peak at about 15-25 keV (depending on colour quenching). The maximum keV is 30-50 (depending on colour quenching). Carbon-14 has a single peak at around 5-10 keV and a maximum keV of about 20-40.

3. I have the added complication that the samples range from water samples to animal tissue to sediment, which means a range of different standards need to be made. I've done this successfully for LSC (liquid scintallation counting) samples before, to compensate for colour quenching, but would the same need to be done for the gamma counter in case there were differences in efficiency? As I understand it, it is the geometry and mass or volume that is most important in the gamma counter. All our samples are quite small (a few g or ml)—do you think it's important to make standards with each of the sample types? I have only ever done this with different sample sizes of the same material (sediment) for the gamma counter before.

The ¹⁴C is actually incorporated as a marker into an organic molecule which should be relatively easy (?!) to extract with a solvent. So that's always an option too.

It sounds as if you are on the right track. Carbon-14 decays with the emission of beta radiation, alone, with a maximum beta energy of 156 keV, which means that the beta spectrum consists of beta particles that range in energy from 0 to 156 keV, the average energy being about 50 keV. Cadmium-109 decays by electron capture to 39 second half-life ¹⁰⁹Ag, which deexcites by isomeric transition to ground state. The radiations emitted in the Cd and Ag decay process include a number of monoenergetic Auger electrons and conversion electrons that range in energy from about 2 keV to 88 keV; additionally, some x rays, lower in energy than 25 keV, and a low yield 88 keV gamma ray are emitted. The end result, from the point of view of simultaneous analysis of ¹⁴C and ¹⁰⁹Cd using LSC is, as you have observed, that the energy deposition events in the scintillation fluid overlap to the point where the two radionuclides cannot be reasonably discriminated.

Your suggestion of counting the sample first in a gamma counter to determine ¹⁰⁹Cd activity and then subtracting this activity from the total activity determined by LSC to obtain the ¹⁴C activity is, indeed, legitimate, but you must take care that you have properly evaluated the efficiency for ¹⁰⁹Cd counting by LSC and by gamma counting. Even though the ¹⁰⁹Cd and the ¹⁴C induced pulses overlap in LSC, the energy spectra are different, and you cannot apply the same efficiencies to both radionuclides. If you still have some of the ¹⁰⁹Cd standard available, you can make up a sample containing a known amount of ¹⁰⁹Cd in a sample matrix as similar as possible to that being evaluated and use this to determine the ¹⁰⁹Cd LSC counting efficiency. In determining this efficiency, do not change any of the settings on the scintillation counting system. (I am assuming that you have already set up the liquid scintillation system for ¹⁴C counting and an appropriate ¹⁴C channel or window has been established.) Treat the counting process for the ¹⁰⁹Cd as if you were counting ¹⁴C. To determine the ¹⁰⁹Cd counting efficiency, count the standard for a preset time, record the net count rate in the ¹⁴C channel, in counts per minute, cpm, and divide the net count rate by the standard activity, in disintegrations per minute, dpm, to obtain the efficiency in units of counts per disintegration, c/d.

You should also prepare a ¹⁰⁹Cd standard for use in the gamma counter to evaluate the counting efficiency, also in c/d, by gamma counting. This standard should be prepared in the same volume and geometry as will be used when the actual sample of interest is gamma-counted. Most of the photons emitted from ¹⁰⁹Cd are x rays with energies less than 25 keV and, depending on the characteristics of your gamma detector, these may not be seen with high efficiency. The single gamma ray of energy 88 keV should be detected with high efficiency, but it has a low yield of 3.6%. Thus, you should make sure you gamma count long enough to get a reasonable number of net counts. In the analysis of the actual sample you would divide the sample ¹⁰⁹Cd net gamma count rate, likely in counts per minute, cpm, from the gamma counter by the determined gamma counting efficiency to obtain the ¹⁰⁹Cd activity in dpm. When the sample is then counted by LSC you would multiply the ¹⁰⁹Cd sample activity from the gamma analysis by the efficiency determined when you ran the ¹⁰⁹Cd standard in the liquid scintillation system to get expected LSC count rate, in cpm, from the ¹⁰⁹Cd, and then subtract this count rate from the total net count rate (call the total net count rate Ctot) in the ¹⁴C channel obtained from the sample LSC count. The subtracted count rate, call it Csub, would be that due to ¹⁴C. This value could be divided by the ¹⁴C counting efficiency, if this is explicitly known, to obtain ¹⁴C activity. Since your liquid scintillation system probably outputs "total ¹⁴C activity" (call this Atot), based in this case on the combined ¹⁴C and the ¹⁰⁹Cd counts, you could also calculate the ¹⁴C activity by multiplying the ratio Csub/Ctot by Atot.

There are a number of chemical methods that could be used for separating the ¹⁴C from the ¹⁰⁹Cd, but these require knowing the chemical forms of the two species and being familiar with some applicable separation schemes, such as ion exchange chromatography or solvent extraction. These are beyond the scope of this discussion.

The standard and sample color and chemical composition are important for LSC as you have acknowledged. Your understanding is also correct that these factors are not particularly important for gamma counting and the sample volume and mass are the most important things. If your sample volumes are small and, especially if you are using a well-type detector in which the sample vial sits in the well, small variations in volume between the sample and standard are not very consequential. If the sample is not in a detector well but sits on or close to the detector surface, you should keep the sample and standard volumes as similar as possible. Also, since the photon energies from ¹⁰⁹Cd are quite low, the sample and standard masses (and preferably atomic number) should be similar so that they present about the same level of photon attenuation. For aqueous liquid samples and soft tissue samples, an aqueous standard of the same volume as the sample(s) should be suitable. If you are using any bone samples and for sediment samples, the mass densities and effective atomic numbers of these materials are significantly different from water, and you may have to attempt to make standards that have similar physical characteristics, more difficult but doable.

It appears that your inclinations so far have been sound, and you should be successful in your endeavors. The few suggestions made above are intended to point out factors, related to detection efficiency, that are important in the analysis. Good luck!

George E. Chabot, PhD, CHP