The ab-initio calculations of the stability of unknown materials present new developments in Solid State Chemistry and can improve the research of new superhard materials. The compound C3N4 is an interesting illustration of such a new route for predicting original materials. During approximately the past 10 years five different structures have been predicted : one is bidimensional (the graphitic form of C3N4) and four are 3-dimensional (alpha, beta, a zinc-blende-type with carbon vacancies, a cubic-type isostructural of the high-pressure variety of Zn2SiO4). Different routes have been investigated for preparing a variety of C3N4 (involving P.V.D., C.V.D. and high pressures processes). Unfortunately, due to the strong stability of the N-2 molecule, carbonitrides deficient in nitrogen have been obtained in most of the cases. In several attempts, small microcrystallites attributed to a crystallized C3N4 form embedded in an amorphous phase have been isolated. Recently using a solvothermal process : the pyrolysis of melamine in presence of a nitriding solvent (NH2NH2) in high pressure conditions (P approximate to 3GPa, 800 less than or equal to T less than or equal to 850 degrees C), the graphitic form of C3N4 have been prepared and characterized (XRD, SEM, Electron probe micro-analysis, ESCA, I.R. spectroscopy).