Limitations of measuring small magnetic signals of samples deposited on a diamagnetic substrate

The superconducting quantum interference device (SQUID) magnetometer is one of the most sensitive experimental techniques to magnetically characterize samples with high sensitivity. Under realistic experimental conditions, especially for samples on a substrate, the magnetic signal as recorded with a commercial MPMS-SQUID magnetometer, is dominated by artifacts below about 4 x 10(-7) emu. Here we present a detailed discussion of possible artifacts and pitfalls characteristic for commercial SQUID magnetometers. We establish a limit below which the presence of a magnetic hysteresis is only necessary but not sufficient to claim the existence of ferromagnetism. (C) 2008 Elsevier B.V. All rights reserved.