Possible Errors in the Ferrous Sulfate Dosimeter

The latest report of the I.C.R.U. (1) states that the ferrous-ferric dosimeter is capable of measuring absorbed dose in amounts greater than 4,000 rads with a precision of ±1 per cent. Another report (2) gives ±2 per cent as the limit of overall reproducibility, but attributes at least some of this variation to the nature and internal dimensions of the ampules which contain the ferrous solutions. In our calibrating procedure for the electron beam of a betatron, we have noticed some discrepancies in results when this dosimeter is employed. Our investigation of these discrepancies forms the basis of this report. Studies disclosed that an error whose magnitude depends on the nature and size of the container may be introduced and further suggested a means for correcting a measurement. Our calibrating procedure made use of a phantom into which Perspex cylinders could be placed in a reproducible fashion. Three cylinders were utilized, one containing an ionization chamber, one holding 5 capsules each containing 60 mg of LiF, and one polystyrene test tube containing 5 ml of ferrous sulfate dosimeter solution. To standardize conditions from one dosimeter to the other as much as possible, a Perspex plug was employed to fill the air space above the solution. An attempt was made to keep the solution from touching the plug but this was not always achieved. The solution was also left standing in the test tube both before and after irradiation for total periods of up to twenty-four hours. As will be seen, although it was not apparent at the time, both these factors are of significance. Our ferrous-ferric dosimeter readings were made by taking two aliquots of solution from the irradiated test tube and obtaining separate ferric ion déterminations from each. A discrepancy was observed between these two readings which increased if the test tube were shaken just before removal of the second sample. This procedure was repeated with unirradiated solution except that the Perspex was not employed, nor was the solution left standing in the test tube for extended periods. The difference between the shaken and unshaken readings was much smaller. Initially, we considered this to mean that irradiation of the solution was involved in the production of the difference in reading, but we subsequently found that irradiation had nothing to do with the effect. An attempt was made to establish the nature of this increased reading by subtracting the absorbance spectrum of the unshaken from that of the shaken solution. The difference was characteristic of scattered light and suggested that the increased absorbance of the shaken solution was due to the presence of a suspension in it. At first the purity of our chemicals was suspected, but we subsequently found no difference in the effect when we substituted analytical grade ferrous sulfate from two other drug houses, water doubly distilled from alkaline permanganate solution,