Magnetic measurements with a SQUID magnetometer: Possible artifacts induced by sample-holder off-centering

Abstract
Asymmetries in sample holders used for magnetic measurements on magnetometers based on induction methods, such as those equipped with SQUID sensors, can lead to substantial errors and/or important artifacts which resemble phase transitions. They occur under the conditions for which sample and/holder have signals of opposite sign, but are nearly equal in magnitude. The most serious errors can occur often, but not exclusively, for compounds having intermediate magnetic dilution. We present here a general discussion of the problem illustrated by measurements of a polyoxometallate with the known Keggin structure K6[Fe(III)W12O40]⋅nH2O, done on a SHE VTS model 905. While one of the obvious solutions to this problem is the use of holders with a signal much smaller than that of the sample, it would be preferable if, in addition, the holder had a response of the same sign as that of the sample, for all temperatures and fields measured.