Positron Shields and Why People Keep Doing Them Wrong
I got called to a cyclotron lab last month because someone had built a lead box around an F-18 source and their pocket dosimeter was reading 80 microsieverts per hour at 30 centimeters. The actual source activity was modest—maybe 2 curies—but the dose rates made it look like they were shielding a cobalt-60 unit. The problem was simple and completely preventable. They had used lead as the first layer. Lead has a high atomic number, which means positrons passing through it dump a ton of energy as bremsstrahlung X-rays. What they should have done is put acrylic on the inside to slow the positrons down before they ever touched the lead. Beta-plus decay produces a positron, which is an electron's antimatter twin. When a positron leaves the nucleus it has kinetic energy determined by the isotope in question. F-18 positrons max out at about 0.635 MeV. That is low energy compared to typical beta-minus emitters. O-15 is around 1.7 MeV. Na-22 sits near 0.545 MeV. The maximum energy matters because it determines how far the positron travels through matter before it stops. The range of positrons in plastic is roughly 0.5 millimeters per MeV of maximum energy. For F-18 that means you need about 3 millimeters of acrylic to stop essentially every positron. For O-15 you are looking at 8 to 10 millimeters. This is not a lot of material. A standard 1-centimeter-thick acrylic sheet will stop F-18 positrons with plenty of margin. The same sheet does nothing useful against the gamma rays that appear right after the positrons stop, and that is where the real shielding work begins.
Positron annihilation is unavoidable. Once a positron loses its kinetic energy it meets an electron and both disappear, producing two gamma photons each with exactly 511 keV of energy traveling in roughly opposite directions. Those 511 keV photons are what make positron sources annoying to shield. They penetrate lead, concrete, and thick acrylic with similar effectiveness. The shielding strategy for beta-plus emitters is therefore two-stage: stop the positron with low-Z material, then attenuate the annihilation photons with high-Z material.
Why Layering Matters More Than Thickness
A single layer of lead might look efficient on paper. You look at a half-value layer chart, see that 511 keV photons get cut in half by about 4 millimeters of lead, and decide to wrap your source in an inch of it. That is a bad decision for anything that emits positrons. The positrons will strike the lead and generate bremsstrahlung. The bremsstrahlung spectrum from a 0.635 MeV positron in lead peaks around 300 to 400 keV and extends up toward the original maximum energy. You have replaced a manageable positron problem with a harder X-ray problem. The dose rate at your working distance can actually increase after you add the lead. I learned this with a Na-22 source that was sitting in a brass holder on a benchtop. Brass is roughly 67 percent copper and 33 percent zinc, both relatively high-Z metals. I measured the dose rate with a Geiger counter positioned near the source and saw readings about four times what I calculated from the gamma constant alone. I suspected bremsstrahlung and tested it by wrapping the source in a piece of thin aluminum foil. The readings dropped to calculated levels immediately. The foil stopped the positrons and eliminated the excess bremsstrahlung from the brass. The fix took thirty seconds and a piece of scrap metal. The correct layering order is always low-Z first, high-Z second. Acrylic or polycarbonate on the inside, lead or tungsten on the outside. For a typical F-18 syringe shield, you will see about 1 centimeter of acrylic liner inside a tungsten body. The acrylic eats the positrons. The tungsten eats the 511 keV gammas. Some vendors sell syringe shields that skip the acrylic liner and use bare tungsten or lead. Those work fine for pure gamma emitters. They are terrible for positron sources.
Get the Full Details
Specific Shielding Calculations for Common PET Isotopes
F-18 is the most common beta-plus emitter in clinical and research labs. Maximum positron energy is 0.635 MeV. CSDA range in water is about 2.4 millimeters. In acrylic it is roughly 2.8 millimeters. A 6-millimeter acrylic barrier gives you a safety factor of more than two. The annihilation radiation from F-18 is two 511 keV photons per decay. The specific gamma ray constant for F-18 is approximately 0.159 millisievert per hour per gigabecquerel at one meter. A 5-gigabecquerel source produces about 0.8 millisievert per hour at one meter with no shielding. Add 6 millimeters of acrylic and nothing changes for the gammas. You then need about 2.5 centimeters of lead to bring that down to roughly 0.1 millisievert per hour, or one half-value layer of lead for 511 keV photons. O-15 has a much higher maximum positron energy at 1.73 MeV. The range in acrylic is about 8 millimeters. You should use at least 10 millimeters of acrylic to be safe. O-15 decays in 122 seconds, so time matters more than shielding thickness for this isotope. If you are working with O-15, the fastest thing you can do is move quickly and keep the source as far from your body as possible. The half-life does the rest. Na-22 is interesting because it is not a PET isotope but it is used for calibration and produces a 1.27 MeV gamma in addition to the annihilation photons. The beta-plus branch is about 90 percent. The range in acrylic is still only about 3 millimeters because the maximum energy is 0.545 MeV. But the 1.27 MeV gamma is harder to shield. You need about 1.2 centimeters of lead for one half-value layer at that energy, compared to 4 millimeters for the 511 keV photons. A Na-22 source demands more lead than an F-18 source of the same activity because of that extra gamma line.
Practical Workarounds That Save Time and Dose
I run a small radiochemistry lab and we process F-18 at activities up to 50 curies during synthesis. The hot cell has lead walls about 5 centimeters thick and an acrylic window. The acrylic window is not optional. Without it the positrons hitting the lead glass would create bremsstrahlung that degrades the image quality on the cameras and increases the dose to anyone standing nearby. We replace the acrylic liner in our syringe shields every three months regardless of damage. The acrylic cracks from gamma exposure over time and thin spots let positrons through to the tungsten. One edge case that cost us a week of troubleshooting involved an F-18 transfer line. We were moving activity through a stainless steel tube and the dose rate at a joint near the wall was 30 microsieverts per hour higher than simulation predicted. We initially blamed a leak. It was not a leak. The positrons were traveling through the steel and producing bremsstrahlung inside the tube wall. We lined the transfer path with a flexible PTFE sleeve and the readings normalized. PTFE is slightly less dense than acrylic but has the same benefit of low atomic number. It cost about twenty dollars and solved a problem we had been chasing for days. For storage, we use lead pigs with removable acrylic liners. The liners are easy to swap and inspect. A lead-only storage container for a positron source is a mistake waiting to happen. The positrons will migrate through any gaps or imperfections and hit the lead. Even a thin layer of acrylic, maybe 3 millimeters, lining the interior of the lead container eliminates the bremsstrahlung problem. I have seen commercial containers sold for beta-plus sources without any inner liner. Do not buy those.
When Shielding Fails Completely
There are scenarios where no amount of shielding solves the problem. High-activity O-15 in a compact geometry will produce significant dose rates even through several centimeters of lead. At those activity levels the primary control is time and distance, not shielding. Keep the source in a shielded container, move it with a mechanical manipulator, and spend only the minimum time near it. A 5-centimeter lead brick between you and an O-15 source will reduce the dose rate substantially, but if you are holding the source by hand the positrons will irradiate your fingers directly before they ever reach the lead. Distance is your best tool here. Another failure mode is improper geometry. Shielding calculations assume a point source in free space. A bulk liquid source in a wide container behaves differently. The self-absorption in the liquid reduces the external positron fluence but the annihilation photons escape from the entire surface area. A flat container of F-18 solution will produce a more uniform dose field than a tall thin vial of the same activity. Your shielding needs to account for the actual shape, not just the activity. I once designed a shield based on a point-source calculation for a 2-liter batch of F-18 in a reactor vessel. The dose rates outside the shield were double what I had estimated because the broad surface area was acting like an extended source. I added another centimeter of lead and reduced the hot cell access time by half. Depleted uranium is sometimes recommended for high-energy beta-plus sources because it is denser than lead and provides more attenuation per centimeter. It is also pyrophoric and regulated. For routine PET work it adds complexity without meaningful benefit. Lead is easier to machine, easier to source, and perfectly adequate when the two-layer design is followed correctly. Tungsten is preferable for syringe shields because of space constraints, but again, only when paired with an inner acrylic liner.
Inspection and Maintenance
Acrylic yellows and crazes under prolonged gamma exposure. Check your liners monthly. If you see stress cracks or cloudiness that is not just surface scratching, replace the liner. Cracked acrylic lets positrons reach the metal underneath and creates bremsstrahlung at the crack sites. I use a simple visual inspection with a bright flashlight held at an angle to the surface. Light scatters differently at crack sites and makes them obvious. Lead deforms over time under its own weight, especially in thick sections. Shield walls more than 10 centimeters thick can sag. Check for gaps that open at the bottom of your shielding assemblies. A gap of even a few millimeters becomes a leak path for 511 keV photons. Seal gaps with additional lead putty or bolt-on lead strips. I have a storage cabinet where the bottom door had developed a 2-millimeter gap from warping. The dose rate at the base of the cabinet was 40 percent higher than the top. A strip of lead tape fixed it. Always map the dose rate around any new shield before declaring it adequate. A survey meter with a thin end-window probe will detect positrons directly if they are escaping. A standard Geiger-Müller tube with a thick wall may only see the gammas and miss a bremsstrahlung problem entirely. Use both types of probes when commissioning a new shield. The difference between the readings tells you whether positrons are getting through.
The fundamental principle is straightforward and rarely controversial among people who actually work with these sources. Positrons stop in plastic. Gamma rays stop in metal. Put the plastic between the source and the metal. Follow that rule and most shielding problems go away on their own. Ignore it and you will spend more time fixing dose rate anomalies than you would have spent doing the actual work.
