In water-bearing porous rocks, pore pressure variations play a major role in deformation, through dissolution−precipitation and fracturing processes. An often-overlooked variation where pressure falls to negative pressure or tension can operate whenever aquifer formations dry out, for instance, in deep storage (nuclear or industrial wastes, long-term CO 2 mitigation, short-term energetic resources, etc.). This can generate capillary tension within the aquifers. This study investigates the mechanical effect of such in-pore tension in the surrounding crystal field, through laboratory experiments at the one-pore scale. Microthermometric procedures were carried out on synthetic fluid inclusions to generate large tensile stress and were combined with Raman microspectrometry to visualize the resulting stress fields in the host quartz. For comparison, we numerically modeled the stress field by linear elasticity theory. The experiments demonstrate that significant damage is produced in crystalline materials by the pore tension. Despite the induced stress measured by micro-Raman spectrometry to remain moderate, it is able to fracture the quartz. The volume of the cavity is a prominent controlling parameter for the stress amplitude. The crystalline heterogeneities of the solid are another major parameter for localizing the mean weak stress and accumulating overstress. Our results call for bringing pore-scale micromechanics into the safety assessment of the geological storage of various wastes inside depleted aquifers. They also show the magnifying effect of heterogeneities on propagating stress and localizing it along certain directions, promoting the final failure of water-bearing minerals, rocks, or pore networks. KEYWORDS: tensile strength, pore rock damage, storage safety, Raman scattering, quartz weakening, fluid inclusions 45 of chemical origin, which contributes to water weakening. 5,6 In 46 particular, imaging and quantification of the microstress in the 47 solid network, before and after fluid tension sets up in the 48 pores, have been missing until now. 49 The challenges associated with this area of research are first 50 in materials science in which several examples of tension-51 driven fracture processes still require mechanistic explanations 52 or quantification. While sol−gel processing explains how 53 drying-driven stress causes bodies to crack, how nanoparticle 54 suspensions dry is much less understood, depending 55 appreciably on molecular-scale physics combining capillary 56 and viscous forces (e.g. , refs. 7,8 ). How cementitious materials 57 dry is also of interest. There is little doubt there, ever since the 58 pioneering work 9,10 showing that the main driving force is