The feasibility of using microfluidic tests to investigate water-rock (mineral) interactions in fractures regarding sorption onto thin rock sections (i.e., shale and granite) of lead (Pb) and uranium (U) was evaluated using a synthetic PbCl2 solution and uranium-containing natural groundwater as fluids. Effluent composition and element distribution on the thin rock sections before and after microfluidic testing were analyzed. Most Pb removal (9.8 mg/cm(2)) occurred within 3.5 h (140 PVF), which was 74% of the total Pb removal (13.2 mg/cm(2)) at the end of testing (14.5 h, 560 PVF). Element composition on the thin shale sections determined by mu-XRF analysis indicated that Pb removal was related primarily to Fe-containing minerals (e.g., pyrite). Two thin granite sections (biotite rich, Bt-R and biotite poor, Bt-P) exhibited no marked difference in uranium removal capacity, but a slightly higher amount of uranium was removed onto the thin Bt-R section (266 mu g/cm(2)) than the thin Bt-P section (240 mu g/cm(2)) within 120 h (4800 PVF). However, uranium could not be detected by micro X-ray fluorescence (mu-XRF) analysis, likely due to the detection limit These results suggest that microfluidic testing on thin rock sections enables quantitative evaluation of rock (mineral)-water interactions at the micro-fracture or pore scale. (C) 2016 Elsevier B.V. All rights reserved.