Answer to Question #11421 Submitted to "Ask the Experts"

We use thermoluminescent dosimeter (TLD700) chips for our in vivo dosimetry. These chips are suddenly showing an increase of approximately 60% in their thermoluminescent sensitivity to 6 megavolt (MV) x rays. The reference light value from our Harshaw 5500 TLD reader has not changed.

Assuming that this change applies to all of the TLDs being used for such measurements and that you are not using a different batch of TLDs or different type of TLD from those previously used, I find it difficult to explain a change as large as 60% in the response of your TLDs to these x rays. I would not expect such a large difference in response unless some notable change had occurred in one or several of the following:

1. The manner in which the TLDs are being handled and processed (e.g., a significant difference in how the TLDs are being annealed or a marked change in the readout cycle parameters).
2. The physical configuration of the TLDs (e.g., changes in any holder being used to hold/position the dosimeter or changes in composition or position of any filtration or scattering materials incorporated into the TLD holder).
3. Changes in positions of the TLDs on the phantom compared to that used in the past for assessment/calibration purposes (e.g., if the TLDs being irradiated are in a bare configuration near the surface of the phantom compared to earlier measurements with the dosimeters at some depth or covered by some filtration).
4. Changes to the intensity and/or energy distribution of x rays from the machine (e.g., possible changes in filtration or collimation materials or configuration or a possible change in the effective voltage).

You state that the reference light source in the TLD reader is yielding the same results as it has previously, thus confirming that the reader output per unit light input appears to be correct. I assume that you have validated the operating parameters and the output of the 6 MV accelerator using a calibrated ion chamber or equivalent device.

The next thing I would do, if you have not already done so, is to irradiate a group of the TLDs in a known configuration (preferably the same as that being used at 6 MV) to a known dose of gamma radiation or x rays at a lower energy, preferably between about 250 kiloelectronvolts (keV) and 1.5 megaelectronvolts (MeV). Cobalt-60 (⁶⁰Co) gamma radiation would be good for this if you have it available. The source should be one for which you have previously evaluated, or know by other means, the expected response of the TLDs. If the results are as expected, and the TLDs continue to yield a 60% excess response to the 6 MV x rays, then we might conclude that some change such as one or more of those noted above may have occurred.

In the event that the TLDs being used have been subjected to some altering influence that has resulted in the overresponse, I would then try irradiating a new group of the same model TLDs that you have been using to both a lower-energy source and then separately to the 6 MV x rays, reading the dosimeters out after each irradiation to evaluate whether the new batch of TLDs is yielding a similar overresponse as the batch you had been using. If the overresponse persists in the new batch, then one would conclude that the cause may be associated with a change that had occurred in the irradiation conditions, possibly related to one or more of the factors noted earlier.

As you likely know, lithium fluoride (LiF) TLDs have an inherent photon dose response that is fairly constant with energy, exhibiting a maximum overresponse, perhaps 40%, at around 30 keV. Thus, if you were using the TLDs to view largely low-energy photons, which might be present as scattered radiation outside the intended field, you would expect some overresponse if the dosimeters had been calibrated at a higher energy but not likely an overresponse that was near 60% high. Such overresponse is often associated with the increased photoelectric effect, which has a 4th to 5th power cross-section dependence on atomic number, Z, and an approximately cubic inverse dependence on energy. The slightly higher atomic number of LiF compared to soft tissue accounts for the excess response at low energies.

Another contributor to excess dose response, relative to tissue, is sometimes observed at high photon energies, greater than 1.022 MeV, when the TLD element is covered by or backed by a high-Z material such as lead. In such cases the elevated response results from electrons and positrons produced by pair production interactions in the high-Z material, with the cross section exhibiting an approximately Z-squared dependence. Such an effect can also contribute to dose from high-Z filters and collimator devices in the x-ray beam as pair production events in the high-Z materials produce electrons and positrons. These charged particles produce excess dose that might be detected if the TLDs are close to the surface of the phantom or body.

It appears that you have some detective work to do to resolve the problem. I hope these suggestions are helpful.

George Chabot, PhD, CHP