Mapping the Neutron Flux Distribution Near a Medical LINAC
Because of the great interest in visiting Mars by NASA, private companies, and governments around the world, we are interested in studying the neutron shielding effects of different materials. Astronauts on a Mars mission will be exposed to space radiation, which might have profound health effects. Shielding astronauts from neutrons is especially difficult because they do not carry a charge.
In order to study shielding effects, we first needed a source of neutrons. We found this in a medical linear accelerator (LINAC) through a collaboration with the medical physics department at Massachusetts General Hospital (MGH). LINACs are used in the radiation treatment of cancer patients, and they produce neutrons as a by-product.
The next step in this research was to map the neutron flux, in neutrons per cm2 per second, around the LINAC. From published simulations of medical LINACs, we know there is a dominant peak in neutron production centered at a high-energy peak of 1 MeV and a low-energy peak around 0.025 eV. We wanted to determine the neutron flux associated with these peaks for this particular machine so that we could calculate the Q value, or neutron production coefficient, as neutrons emitted per Gy. (Gy is the SI unit for an ionizing radiation dose.) We could then compare that number with published values for other 15 MV machines.
First, we investigated the high-energy neutron flux using bubble detectors—an easy and inexpensive method of detecting neutrons above 200 keV. Bubble detectors contain a gel that is packed with microbubbles of organic fluid. When energy is transferred from an incoming neutron to a microbubble, the microbubble expands and forms a visible bubble a few millimeters in diameter. The detectors are calibrated so that the number of bubbles after irradiation corresponds to the fluence of neutrons to which the detector was exposed.
For our experiments, we positioned the detectors in a straight line emanating radially out from the head of the LINAC. After radiating the detectors, we analyzed the relationship between fluence and distance from the head. (We also made distance measurements at different angles with respect to the LINAC head but determined there was no significant angular dependence.) Then we calculated the neutrons per area for a given radius and integrated over 4πr2 to obtain the Q value for the machine in neutrons/Gy. Our results indicate a high energy flux of 106 neutrons/cm2/Gy, which is on the same order of magnitude as published values for 15 MV machines.
While bubble detectors are useful tools for neutron detection, they have limitations. For instance, the calibration is only accurate to 20 percent. The bubble detectors can be reused by means of a built-in plunger that recompresses the bubbles, but after several cycles we observed that sometimes the recompression failed to eliminate all the bubbles. Additionally, bubble detectors have a threshold energy limit of 200 keV.
To determine the low-energy neutron fluence, we performed a neutron activation analysis using thin metal foils with high purity. The process works this way. When neutrons bombard a thin, pure metal foil, they create radioactive isotopes. By studying the radioactive emissions and decays from a foil, you can work backward to determine the amount of incident neutrons.
For this analysis to work in practice, you need a metal that reacts with neutrons in your specific energy range of interest and that produces an isotope with a reasonable half-life. For us, this meant the half-life must be short enough to achieve high activity and good statistics following an hour-long irradiation by the LINAC, but long enough so that the activity did not decrease too much before reaching our detector. As a detector we used a high-efficiency beta detector that belongs to MGH, so we also needed the isotope to emit beta rays when it decayed. Based on these considerations, we chose copper, indium, and gold foils 0.5 inch in diameter and 0.002–0.01 inch thick.
Analyzing the foils was an interesting challenge. Some of the beta rays were shielded internally by the foils, depending on their energy, point of origin within the foil, and trajectory outward. In addition, a given isotope has a well-defined maximum energy for the beta, but there was a distribution of energies. For these reasons, we needed a correction factor to estimate the total number of betas emitted based on the number that escaped. This correction factor depends on the foil.
Our results indicate a low-energy neutron flux on the order of 104 neutrons/cm2/Gy, which is in the range of published values for similar machines. The next step in this process is to continue foil irradiation and analysis using a variety of materials, both shielded with cadnium covers to absorb thermal neutrons and unshielded to create a full map of the neutron flux around the LINAC. This will allow us to eventually use the LINAC to test a variety of materials for their durability during long-distance space missions where high levels of radiation will be absorbed.
Through this process, we have learned a variety of new skills, from teamwork and leadership to innovation, perseverance, and analysis. This opportunity has helped us grow academically and allowed us to explore areas of physics we may not have been exposed to otherwise.
Acknowledgments: The Suffolk University team includes physics majors Paul Johnson, Mario Rojas, Allen Alfadel, John Thomas, Jackson Nolan, Erick Bergstrom, Molly McDonough, Brian Hassett, and Dylan Barbagallo, along with faculty research advisor Dr. Walter Johnson and educational coordinator for medical dosimetry Jacky Nyamwanda. The LINAC is from the Massachusetts General Hospital Oncology Department, and our collaborators are Dr. David Gierga and Tara Medich. This research project would not have been possible without a 2017–18 SPS Chapter Research Award.
