Abstract
Magnetic pore fabrics (MPF) are an efficient way to characterize pore space anisotropy, i.e., the average pore shape and orientation. They are determined by impregnating rocks with ferrofluid and then measuring their magnetic anisotropy. Obtaining even impregnation of the entire pore space is key for reliable results, and a major challenge in MPF studies. Here, impregnation efficiency and its spatial variability are systematically tested for natural (wood, rock) and synthetic (gel) samples, using oil- and water-based ferrofluids, and comparing various impregnation methods: percolation, standard vacuum impregnation, flowthrough vacuum impregnation, immersion, diffusion, and diffusion assisted by magnetic forcing. Seemingly best impregnation was achieved by standard vacuum impregnation and oil-based ferrofluid (76%), and percolation (53%) on rock samples; however, sub-sampling revealed inhomogeneous distribution of the fluid within the samples. Flowthrough vacuum impregnation yielded slightly lower bulk impregnation efficiencies, but more homogeneous distribution of the fluid. Magnetically assisted diffusion led to faster impregnation in gel samples, but appeared to be hindered in rocks by particle aggregation. This suggests that processes other than the mechanical transport of nanoparticles in the pore space need to be taken into account, including potential interactions between the ferrofluid and rock, particle aggregation and filtering. Our results indicate that bulk measurements are not sufficient to assess impregnation efficiency. Since spatial variation of impregnation efficiency may affect MPF orientation, degree and shape, impregnation efficiency should be tested on sub-samples prior to MPF interpretation. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s11242-022-01809-0.