In this paper we establish that the yield of secondary positive monoatomic ions of silicate targets is a linear function of the cationic concentrations. This law can be applied for quantitative analysis of silicates, minerals and glasses, in situations which were practically out of scope until now : small phases (down to 0.5 μm) or trace concentrations (5 x 10-3 to 10-6). It represents a linear approximation which can be qualitatively understood by a bond breaking model