The use of renewable lignocellulosic resources experiences an undeniable resurgence of interest for building materials development. Based on this observation, a first line of work has seen the development of plant-based concretes constituted of lignocellulosic aggregates and a pozzolanic binder. Hemp and sunflower stem aggregates were then selected for their similarities and availability in the Auvergne region. In parallel, pozzolanic binder consisting of 80% by mass of volcanic pumice sand from a local quarry, 20% lime and an activator has been formulated. The results show the similarity of hemp and sunflower aggregates characteristics as well as mechanical and thermal properties of lightweight concretes (<500kg.m-3) made of these plants. Formulated materials meet the criteria appointed in French professional rules of construction in hemp. In a second line of work, the literature review driven enabled to identify interface between the plant particles and the inorganic binder as one of the major scientific issues concerning concrete plants. Heart hardening problems of these materials indeed deserve to be linked with those observed at the interface binder/wood. Two ways of improvement are explored in a multi-scale approach combining physico-chemical analysis and mechanical testing: pre-treatment of vegetable particles and specific adjuvantation of the binder. The first strategy was to make on the plant aggregates two types of changes: a linseed oil recovery and a treatment in an aqueous solution of Ca(OH)2. In order to analyse the short-term interaction between the modified particles and the binder, a specific measurement device based on an immersed plate is used. The results show an improvement of the binder/plant interphase after aggregates treatment in Ca(OH)2 solution. Surprisingly, the plant concretes mechanicals performances fall when plants are pre-treated. This approach is finally restrictive for an industrial point of view and does not solve satisfactorily binder/wood interface problem. The second implemented strategy required the adding in the binder of a water- retaining agent (cellulose ether) in amounts ranging from 0.5 to 1.5% based on the weight of the powders. Observations of samples performed on a SEM microscope and coupled with an EDX analysis underline the ability of the cellulose ether (CE) to improve binder/wood interfaces. An explanation of the action of CE molecules on polymer interfaces based on the literature and analysis of the binder is proposed. This approach has allowed the development of plant concretes immediately demouldable. Designed concretes also show higher compactability and cohesiveness as well as improved mechanical properties compared to the solutions in use. It should be noticed that this improvement is not made at the expense of thermal properties. CE used in plant concretes offer development opportunities: prefabricated blocks, lightweight coatings.
