Abstract

We developed a scanning superconducting quantum interference device (SQUID) microscope for imaging the magnetic field of geological samples in 2015 and have been continuously improving it. Since its introduction, the scanning SQUID microscope system has been used to measure various geological samples, including marine ferromanganese crusts, ferromanganese nodules, single-grain zircon crystals with magnetite inclusions, fault gauge samples, stalagmites with ultra-low magnetizations, ultramafic rocks containing strong magnetic minerals. In this paper, we detail the improvements made to the system and developments of measurement and post-processing software. We improved the implementation and the electrical connection of a SQUID chip on a sapphire rod. A new electrical connection combines aluminum wire bonding and silver paste, ensuring stable and low contact resistance across cooling cycles leading to constant sensitivity and stability of measurements. For the improved system, the shortest sensor-to-sample distance was measured at approximately 125 µm using a precision line current. Electromagnetic shielding and improvements on grounding were made, which contributed to improved signal-to-noise ratios. The introduction of the reference sensor allows subtraction of environmental magnetic field. The new feature of stacking at each measurement grid is effective in reducing noise for weakly magnetized samples. Additionally, the post-processing software has been developed. Various drift correction algorithms provide flexibility to the operator to deal with the situation including a narrow marginal area or a dirty marginal area contaminated with magnetic materials. Spike noises could be removed by median filter, and flux jumps could be corrected by another independent algorithm. Dipole fitting allows the operator to calculate the position, intensity, and direction of a dipole one by one. A list of dipoles is made, which could be subtracted from the magnetic images to clean up for further analyses. Upward continuation enables leveling of the magnetic images measured with different sensor-to-sample distances allowing meaningful comparison and subtraction between different demagnetization steps. Calculation of X- and Y-components from Z-component allows to obtain total magnetic field image, which could be used to locate magnetized materials easily.