[en] The capture and the storage of CO₂ (CSC) on the underground geological formations is a solution for the CO₂ undesirable effects reduction. The geological formations being composed by heterogenic material have a crack network. The fracture toughness (KC) is a rock parameter connected with the capacity of a material to resist crack propagation. The propagation of a crack can be due to the changing of the stress or to the rock degradation. The evaluation of the fracture toughness and its evolution due to the chemical effects is then important to the modelling of the crack propagations on the CO₂ storage context. One objective of this work is the experimental evaluation of the CO₂ degradation on the fracture toughness. For the experimental program a preliminary study was made to the choice of the rock and the test to be performed. In this way, a limestone (Pierre de Lens) was chosen to be studied in an intact and degraded state. The degradation takes place in an autoclave, where the samples are put at CO₂-saturated water in reservoir conditions (60 deg. C and 15 MPa). Several experimental tests are chosen for the evaluation of the fracture toughness in modes I and II. Some examples are the Central crack Brazilian disc (CCBD) for the mode I and the Punch through shear test (PTST) for the mode II. Some tests were performed using an image correlation technique (DIC). This set-up allows the fracture toughness evaluation by the displacement field evolution. The experimental results shows the fracture toughness values obtained by the different configurations are in good agreement. After the DIC technique analysis we can point the fracture toughness in mode II cannot be evaluated in a non-confined test. For the degradation process, the complementary analysis in the mercury porosimeter shows the rock porosity changing is low (0.4 %). At the SEM, the degradation can be observed by a homogenization of the sample. For the fracture toughness it value passes from 0.62 to 0.58 MPa.m0.5. We also have studied the effect of water performing tests in a saturated environment. The influence is more significant with a reduction on 17 % in the fracture toughness. The CO₂ presence in the fluid does not affect this value. Concerning the mode II evaluation by the PTST test, the samples were submitted to different confining pressures (5, 10 and 15 MPa). We can observe a good evaluation of the fracture toughness in mode II (around 3 MPa.m0.5). Nevertheless, the mode II is still present for the pressures 5 and 10 MPa, and it is not inexistent for a 15 MPa pressure. Still in this case the influence of CO₂ is low with values passing from 2.96 to 2.77 MPa.m0.5. The influence if the rock degradation by the CO₂ presence on the crack propagation were studied by the help of a numerical model ENDO-HETEROGENE, that is present in the Code-Aster^R calculation code. This model is based on the initiation and propagation of cracks in a heterogenic environmental where the parameter variability follows the 2 parameters Weibull probabilistic model (m and σ0). We exploited the possibility of the chemical degradation influence on the microstructure heterogeneity that is represented by the parameter m. The model shows that the changing of m influence on the crack number and dimensions by the maximum size of the crack didn't change. Putting this result in context, the Weibull parameters were evaluated for the intact and degraded rock. We observed that m changes from 8.55 to 8.52 and σ0 from 2.8 to 2.2 MPa. The numerical simulations show this variation is not enough to change the crack network that is formed after a load in a geological layer. The general results show that for a limestone reservoir the CO₂ injection affect significantly neither the fracture toughness nor the probabilistic parameters. These results correspond to a 10 years period for a zone far from the injection point. (author)