CONSIDERATIONS ON THE CALIBRATION OF SMALL THERMOLUMINESCENT DOSIMETERS USED FOR MEASUREMENT OF BETA-PARTICLE ABSORBED DOSES IN LIQUID ENVIRONMENTS

An investigation has been carried out on the factors which affect the absolute calibration of thermoluminescent dosimeters (TLDs) used in be ta particle absorbed dose evaluations. Four effects on light output (L O) were considered: decay of detector sensitivity with time, finite TL D volume, dose linearity, and energy dependence. Most important of the se was the decay of LO with time in culture medium, muscle tissue, and gels. This permanent loss of sensitivity was as large as an order of magnitude over a 21-day interval for the nominally 20-mum-thick disc-s haped CaSO4(Dy) TLDs in gel. Associated leaching of the dosimeter crys tals out of the Teflon matrix was observed using scanning electron mic roscopy. Large channels leading from the outside environment into the TLDs were identified using SEM images. A possibility of batch dependen ce of fading was indicated. The second most important effect was the a pparent reduction of light output due to finite size and increased spe cific gravity of the dosimeter (volume effect). We estimated this term by calculations as 10% in standard ''mini'' rods for beta particles f rom Y-90, but nearly a factor of 3 for I-131 beta particles in the sam e geometry. No significant nonlinearity of the log (light output) with log (absorbed dose) over the range 0.05-20.00 Gy was discovered. Ener gy dependence of the LO was found to be not detectable, within measure ment errors, over the range of 0.60-6.0 MeV mean energy electrons. Wit h careful understanding of these effects, calibration via gel phantom would appear to be an acceptable strategy for mini TLDs used in beta a bsorbed dose evaluations in media. Application of these TLDs to monocl onal antibody therapy was of particular interest in these studies.