Investigation of Amorphous Silicon Nanoparticles dispersed in Silicon Nitride Passivation Layer for Semiconductor Industry

During the formation process of a typical integrated circuit (IC), a passivation layer is deposited to protect the internal semiconductor device after the metallization steps. These passivation layers are typically formed by oxide or nitride layers, most of the time deposited by plasma enhanced chemical vapor deposition (PECVD). As part of my PhD research, in collaboration between Politecnico di Milano and ST Microelectronics, I had to perform extensive characterization of different types of silicon nitride (SiN) in order to optimize its adhesion with an epoxy resin deposited on top of it. During these analysis, I detected the presence of amorphous silicon clusters dispersed within the nitride matrix through the presence of strong photoluminescence (PL) effects observed by Raman spectroscopy characterizations. Photoluminescence effects generally originate from the presence of defects or inhomogeneities of such size as to generate energy quantization. In both cases, electrons in the material, excited by the Raman laser beam, occupy energy levels intermediate to the energy gap and subsequently decay emitting photons that result in the formation of very intense bands in the Raman spectrum, called photoluminescence bands. This phenomenon was most pronounced in SiN samples with higher silicon overstoichiometry, which was intentional by company recipe and confirmed by XRD analysis. The silicon overstoichiometry thus led to spontaneous segregation of amorphous silicon into nanoparticles dispersed in the silicon nitride matrix and this phenomenon is almost nonexistent in samples with stoichiometries closer to equilibrium (%At Si/N equal to 0.75). To corroborate the assumptions made from the photoluminescence phenomenon, it was possible to directly observe the nanoparticles by TEM microscopy in some of the characterized samples.<br/><br/>The presence of these nanoparticles is not only interesting from an academic point of view, but could also have an interesting application aspect: one of the major criticalities of passivation layers is their high fragility, and the presence of inhomogeneous agglomerates could lead to an increase in the toughness of the material due to <i>crack deflection</i> phenomena, which have been studied for years in the field of composite materials. This effect occurs when the crack sliding is delayed by the presence of areas with a different thermal expansion index from that of the matrix: at the interface between these areas occurs the formation of compressive or tensile stress fields which attract or repel the crack tip, forcing it to deflect its path and thus increasing the energy required to slide (hence the toughness of the material). The phenomenon of<i> crack deflection</i> is generally observed in composite materials in which the inhomogeneity has a significant size, comparable to or larger than that of the crack tip. In the case of “Silicon Rich” silicon nitride (the sample with high silicon overstechiometry), the nanoparticles observed are too small in size to make an effective contribution to the toughness of the material, and for this reason my project has now focused on finding deposition methods that can lead to the formation of larger amorphous silicon agglomerates.<br/><br/>Obtaining a material suitable for the role of a passivation layer that at the same time exhibits high toughness would be a great step forward in the semiconductor manufacturing industry and could solve several crack failure issues that usually form large blocking point in the development of new devices.