Measurement of radiation from potassium chloride (KCl)

Ever since writing a comment on a thread about how radioactive bananas are, I have been curious about how difficult it would be to detect the radioactivity of potassium chloride (KCl) with a Geiger counter. It turns out that it is very easy. So I thought some people out there might be interested in how I determined how radioactive it is and what kind of radiation it emits.

All potassium has a naturally-occurring radioactive isotope, Potassium-40 (K-40). It is a primordial isotope, meaning that it was formed a few billion years ago through the various stellar processes that created most other isotopes. Since it has a 1.277 billion year half-life, it is long-lived enough to survive to the present day, but there is only one K-40 nucleus out of every 8,500 potassium nuclei (most potassium is K-39 (93%) and K-41 (7%)). Even though K-40 is so rare, this is balanced out by the high abundance of potassium in the world — potassium is found in high amounts in minerals, soil, and organisms. In fact, K-40 gives us an estimated 7% of our total background radiation dose.

Potassium chloride is available at most grocery stores as a salt substitute for people who need to decrease their intake of sodium. It is very cheap — I was able to purchase a 300 gram canister of KCl for $5 at a supermarket. I was previously familiar with its radioactivity, since I have used it in the past to "boost" the K-40 gamma-ray line in germanium detectors to aid in calibration. But I have never tried to measure it with a handheld detector.

For these measurements, I used our lab's survey counter, a Ludlum Model 2241-3 with a Model 44-9 pancake probe sensitive to alphas, betas, and gammas. For the setup, I mounted the probe as shown in the pictures below and did not move it. For a background, I measured an empty glass dish. The glass dish was able to slide underneath the probe, with about 5 mm of clearance.

Next, I placed about 200 grams of KCl in the glass dish (see photo below). According to my calculations, a 200 gram canister should have an activity of 3,200 decays per second (Becquerels) due to K-40. Most of these decays will not be detectable, due to various factors including energy and geometry. But this means that there can potentially be a large increase in counts.

Indeed, the KCl caused the detector's clicking to increase noticeably, as shown here:

Next I set about figuring out what kind of radiation was coming from the potassium chloride. The three kinds of radiation are shielded in different ways. If radiation is primarily alpha particles, a sheet of paper should stop them. So I tried it:

Since the radiation is not attenuated very much (94% passes through), we can conclude that the radiation is not primarily alpha particles.

If radiation is primarily beta particles, then different thicknesses of metal sheets should lead to more and more attenuation. Also, thin sheets of metal will not attenuate gamma-rays very much. Here are my data for some aluminum foil and copper sheets:

The last column is a rough approximation of the energy losses of beta particles passing through that thickness of metal. We can easily see that these data are consistent with a source that emits primarily beta radiation, where the betas have an energy in the hundreds of keV. This description exactly matches the decay data for Potassium-40 (K-40): it emits no alphas, 11% of decays emit high-energy gamma-rays that would be hard to detect with the pancake probe, and 89% of decays lead to a beta particle (endpoint 1311 keV, mean energy 560 keV).

Incidentally, potassium iodide should also be radioactive, though only about 45% as radioactive as potassium chloride by weight.

Mark [BRAWM Team Member]

Background Test	KCl measurement
Paper obstruction	Copper sheet obstruction