Abstract / Description of output
Etch selectivity between layers is an important parameter in the fabrication of microelectronics and microsystems. This is particularly true in the case of isotropic gas/vapour etching methods used to release free standing structures through the selective etching of sacrificial layers. Commonly used structural materials have been reported to be largely inert when exposed to a given vapour etchant, indicating high selectivity when measured against typical sacrificial layers. However, there is growing evidence that these structural layers are actually etched at an enhanced rate if they are located in the proximity of the sacrificial layer being removed. Hence, removal rates given in the literature, which have resulted from measurements of layers that have been etched in isolation, can no longer be trusted to characterize critical etch processes in device fabrication. In this paper, a test structure is reported that enables a far more appropriate determination of the etch selectivity between sacrificial and structural materials. The method is demonstrated with the two most common vapour etch processes. Firstly, the XeF2 vapour etch of a polysilicon sacrificial layer located above a silicon nitride structural layer, and secondly, the HF vapour etch of silicon dioxide placed above a silicon nitride structural layer. Both test structure datasets are presented. The polysilicon and silicon nitride layers, etched with XeF2 show a selectivity of 5:4. The silicon dioxide and silicon nitride layers etched with HF, show a selectivity of 6: 1 to 8: 1.