Clustering solids are a group of materials, which are composed of clathrate compounds [1], semiconductor boron and icosahedral boron-rich solids like boron carbides, dodecaborides, and carbone nanostructures. The atomic structures are complicated and structural defects play an important role, and therefore the control of the electronic valence of these materials is considered to be difficult. Our intention is to overcome this difficulty by gathering, in a unique combination, specialists of theoretical approaches for materials science together with some experimental specialists of the traditional approach of semiconductor technology and of the high-pressure technique [2,3].
In the case of boron for instance, a new idea based on the concept of frustration [4] has been developed to explain the phase stability, and may provide a significant insight into material synthesis. Such a development could be applied to clathrate compounds. Theoretical study bridging different classes of materials will convey a new idea about one material to the other.
Understanding the role of defects in materials such as boron carbides has also proved to be important [5]. This knowledge drives an important concept for the design of materials, and enables us to propose to experimentalists the creation of new materials with improved mechanical properties, provided the valence modification brought in by defects can be controlled.
In addition, the unique thermal properties of clathrates has so far been developed for individual materials. This is a good time to combine these individual properties to establish a unified understanding and relate the thermal properties to the electronic ones.
Finally, the mechanism for the superconductivity in clathrates has been clarified, whereas in boron-based materials the prediction of superconducting temperature is not achieved by the traditional theoretical approach.

This workshop proposal is thus a vehicle to discuss these topics. Holding an international workshop at CECAM in 2013 must increase the research speed and communication of experimental and theoretical communities, allowing progress in the ab initio modeling of clustering materials.