The objective of the Piegeachlor project is to reduce the availability of several families of organochlorine contaminants found in soils. The selected contaminants are polychlorinated biphenyls, polychlorinated dibenzo-dioxins and furans, as well as chlordecone. The initial hypothesis of the project was that carbonaceous matrices resulting from pyrolysis of woody plants such as biochars could play a role of sequestration for this purpose. Following preliminary tests, it turned out that these matrices had on the one hand rather similar characteristics and this in spite of the mobilization of ligneous of different origin and the use of 2 temperatures of pyrolysis (500 and 700°C) and on the other hand a very limited capacity of retention. New carbonaceous matrices were then introduced, in particular activated carbons. The latter could not be produced by the industrial partner and were purchased commercially but characterized with the same rigor as the initial biochars. The sequestration properties were tested using several methodologies. The first one is based on an in vitro approach, qualifying the environmental availability. It simulates the capacity of an aqueous medium to desorb pollutants. The others used a target biota (worm, radish, zucchini, chicken) to determine the relative bioavailability of each contaminant according to the carbon matrix used. These approaches were first applied to OECD-type artificial soils with peat as endogenous organic matter. After a selection of the best performing carbonaceous matrices in terms of sequestration, part of the previously described tests were applied on natural soils collected in France (Saint-Cyprien) and in Martinique (Morne-Rouge, Trinité), respectively contaminated with PCB+PCDD/F and CLD. On the artificial soils, safety tests were conducted on the microbiota and earthworms. The project mainly shows an interaction between the type of test (environmental availability or relative bioavailability depending on the target organism), the carbon matrix used (10 were used at least once) and the contaminants. For the latter, the effect goes beyond the family to which they belong, particularly within the PCBs. It has been shown that the structure of the molecule, through a gradient of polarity or steric hindrance, could be at the origin of very different behaviors. The PCDD/F family is the one that responded best, at least in hens. For carbonaceous matrices, it is clear that microporosity (abundance of pores of the order of nm in size) is a necessary condition for good sequestration of contaminants. It is mainly this property that explains the difference in efficiency between biochars and activated carbons. Finally, a bibliographic study has shown that the choice of highly condensed carbonaceous matrices of the activated carbon type corresponds to a form of carbon that is recalcitrant to biodegradation in soils, which means both an extended capacity over time to sequester the contaminants of interest and also a long term carbon storage. If the economic aspects of this strategy have not been investigated, it should be stressed that this environmental service should be deducted from the overall cost. In terms of perspective, a development of adapted tests for the evaluation of the innocuity is necessary because the carbon matrix is suspected to interact with the good progress of the classic enzymatic tests and thus to bias the results. Similarly, the incorporation rate of the carbonaceous matrix was not tested in the project because of the multiplication of the modalities, but now that certain matrices have been identified for their sequestering power, it would be appropriate to try to optimize it, or even to reduce it in order to limit the deleterious effects on soil life