The PTW microSilicon diode: Performance in small 6 and 15 MV photon fields and utility of density compensation

Purpose We have experimentally and computationally characterized the PTW microSilicon 60023-type diode's performance in 6 and 15 MV photon fields >= 5 x 5 mm(2) projected to isocenter. We tested the detector on- and off-axis at 5 and 15 cm depths in water, and investigated whether its response could be improved by including within it a thin airgap. Methods Experimentally, detector readings were taken in fields generated by a Varian TrueBeam linac and compared with doses-to-water measured using Gafchromic film and ionization chambers. An unmodified 60023-type diode was tested along with detectors modified to include 0.6, 0.8, and 1.0 mm thick airgaps. Computationally, doses absorbed by water and detectors' sensitive volumes were calculated using the EGSnrc/BEAMnrc Monte Carlo radiation transport code. Detector response was characterized using kQclin,4cmfclin,4cm, a factor that corrects for differences in the ratio of dose-to-water to detector reading between small fields and the reference condition, in this study 5 cm deep on-axis in a 4 x 4 cm(2) field. Results The greatest errors in measurements of small field doses made using uncorrected readings from the unmodified 60023-type detector were over-responses of 2.6% +/- 0.5% and 5.3% +/- 2.0% determined computationally and experimentally, relative to the reading-per-dose in the reference field. Corresponding largest errors for the earlier 60017-type detector were 11.9% +/- 0.6% and 11.7% +/- 1.4% over-responses. Adding even the thinnest, 0.6 mm, airgap to the 60023-type detector over-corrected it, leading to under-responses of up to 4.8% +/- 0.6% and 5.0% +/- 1.8% determined computationally and experimentally. Further, Monte Carlo calculations indicate that a detector with a 0.3 mm airgap would read correctly to within 1.3% on-axis. The ratio of doses at 15 and 5 cm depths in water in a 6 MV 4 x 4 cm(2) field was measured more accurately using the unmodified 60023-type detector than using the 60017-type detector, and was within 0.3% of the ratio measured using an ion chamber. The 60023-type diode's sensitivity also varied negligibly as dose-rate was reduced from 13 to 4 Gy min(-1) by decreasing the linac pulse repetition frequency, whereas the sensitivity of the 60017-type detector fell by 1.5%. Conclusions The 60023-type detector performed well in small fields across a wide range of beam energies, field sizes, depths, and off-axis positions. Its response can potentially be further improved by adding a thin, 0.3 mm, airgap.