High-quality data acquisition for first-order reversal curve distributions: Reduction of noise and drift for alternating gradient magnetometer measurements

First-order reversal curve (FORC) distributions are a powerful diagnostic tool for characterizing and quantifying magnetization processes in fine magnetic particle systems, including magnetic minerals. Estimation of FORC distributions is based on calculation of the second-order mixed derivative of magnetic hysteresis data. Diagnosis based on high-quality FORC distributions is important for characterizing weakly magnetized samples, and/or samples with complex mixtures of magnetic minerals. The quality of FORC distributions is controlled by measurement noise and drift. Here we present measurements aimed at characterizing and reducing noise and drift for an alternating gradient magnetometer (AGM) at the Geological Survey of Japan (GSJ). The AGM operates within the range of audible frequencies (typically 250-500 Hz) and the magnetic moment of samples is detected with a piezoelectric sensor while resonated by a set of gradient coils. This means that the AGM is highly sensitive to acoustic and ambient noise, including human voices, instrument buzz, road noise, and building vibrations. The AGM sensors and magnets at GSJ are located within a purpose-built soundproof box, which is placed within a small soundproof enclosure. The controllers and power supplies with cooling fans are placed outside the soundproof enclosure. These measures reduce acoustic noise and improve FORC measurements. In general, drift can be corrected during post-processing (e.g., with FORCinel) with a series of repeated measurements at a constant calibration field that are made during FORC measurements. However, drift cannot be corrected completely if non-linear drift dominates. We find that temperature change is a major source of drift for some FORC measurements at high magnetic fields at GSJ, particularly for weakly magnetized samples with large paramagnetic contributions. Drift was reduced significantly by using a chiller with high precision temperature control for cooling the AGM magnet. Sources of temperature drift and data quality after measurement improvements will be discussed.