We present precise measurements of the resistance-temperature variation of several samples of neutron transmutation doped (NTD) germanium, at temperatures from 70 mK to 1 K. This material is used widely both for thermometry and for the thermistor element in bolometers and microcalorimeters. The resistance, R, is expected to follow the variable range hopping equation R(T) = R_0T~q exp(T_0/T)~p, where T is temperature and R_0 and T_0 are material parameters. A value of ?= 0.5 is supported by theory, and usually appears to allow good fits to data (with the T~q term neglected). However, we find that setting p = 0.5 produces clear systematic errors for some of our samples. Taking p as a fitting parameter gives excellent fits over a large temperature range with p approx= 0.55 for these samples. We consider possible causes for this behaviour, and suggest that in general NTD germanium calibration data should be examined carefully for errors caused by assuming an incorrect value of p.
展开▼
