浸出(土壤学)
多孔性
溶解度
环境扫描电子显微镜
化学
放松(心理学)
松弛法
多孔介质
脉冲场梯度
沥青
材料科学
化学工程
扫描电子显微镜
分析化学(期刊)
复合材料
土壤科学
色谱法
有机化学
自旋回波
磁共振成像
土壤水分
地质学
分子
医学
放射科
社会心理学
心理学
工程类
作者
Rémi Blinder,J.B. Champenois,A. Leclerc,Arnaud Poulesquen,A. Guillermo,Joseph Lautru,Renaud Podor,Michel Bardet
标识
DOI:10.1021/acs.jpcc.9b09285
摘要
We investigate water infiltration in porous matrices made of bitumen and salts with different solubilities. Dispersion, inside bitumen, of either SrSO4 or MgSO4 at 40% mass fraction was achieved, as a way to model materials used in the context of nuclear waste conditioning. Pulsed field gradient nuclear magnetic resonance (PFG-NMR) and environmental scanning electron microscopy (ESEM) measurements allow us to characterize the evolution of the porous structure as a consequence of water infiltration (due to a leaching phenomenon) for durations up to 1.5 year. PFG-NMR enables performing 1D-imaging of water at different times to monitor its slow seeping inside the material. Profiles observed at the maximum time interval (1.5 year) demonstrate that different leaching behaviors exist depending on whether the salts dispersed inside bitumen are of the soluble or insoluble type. NMR relaxation and diffusion measurements were also taken. Remarkably, when performed in combination with 1D-imaging, such measurements yield information on the surface-to-volume ratio of the water-filled porous network, at different times, as a function of depth. In the case of the matrix-containing insoluble salts, relaxation measurements lead to discrimination between two different water populations differing by their T2 or T1 values. These values are described in the framework of surface-driven relaxation. A two-step model of leaching is proposed that globally accounts for the different observations (NMR 1D-imaging, relaxation/diffusion, as well as ESEM). The present work gives insight into the leaching behavior of porous bitumen-salt matrices and could be used as an input for modeling their evolution on longer timescales.
科研通智能强力驱动
Strongly Powered by AbleSci AI