A radiochemical lab-on-a-chip paired with computer vision to unlock the crystallization kinetics of (Ba,Ra)SO4.

(Ra,Ba)SO4 solid solutions are commonly encountered as problematic scales in subsurface energy-related applications, e.g., geothermal systems, hydraulic fracturing, conventional oil and gas, etc. Despite its relevance, its crystallization kinetics were never determined because of radium (226), high radioactivity (3.7 × 1010 Bq g-1), and utilization in contemporary research, therefore constrained to trace amounts (< 10-8 M) with the composition of BaxRa1-xSO4 commonly restricted to x > 0.99. What if lab-on-a-chip technology could create new opportunities, enabling the study of highly radioactive radium beyond traces to access new information? In this work, we developed a lab-on-a-chip experiment paired with computer vision to evaluate the crystal growth rate of (Ba,Ra)SO4 solid solutions. The computer vision algorithm enhances experimental throughput, yielding robust statistical insights and further advancing the efficiency of such experiments. The 3D analysis results of the precipitated crystals using confocal Raman spectroscopy suggested that {210} faces grew twice as fast as {001} faces, mirroring a common observation reported for pure barite. The crystal growth rate of (Ba0.5Ra0.5)SO4 follows a second-order reaction with a kinetic constant equal to (1.23 ± 0.09) × 10-10 mol m-2 s-1.

Microfluidic investigation of pore-size dependency of barite nucleation.

The understanding and prediction of mineral precipitation processes in porous media are relevant for various energy-related subsurface applications. While it is well known that thermodynamic effects can inhibit crystallization in pores with sizes <0.1 µm, the retarded observation of mineral precipitation as function of pore size is less explored. Using barite as an example and based on a series of microfluidic experiments with well-defined pore sizes and shapes, we show that retardation of observation of barite crystallite can already start in pores of 1 µm size, with the probability of nucleation scaling with the pore volume. In general, it can be expected that mineralization occurs preferentially in larger pores in rock matrices, but other parameters such as the exchange of the fluids with respect to reaction time, as well as shape, roughness, and surface functional properties of the pores may affect the crystallization process which can reverse this trend.