The measured lateral stability of polysilicon line structures on a Silicon Oxy-Nitride layer is presented. This has been measured by Lateral Force Microscopy (LFM) in order to understand how much force can be applied to the structure during a wet cleaning process with subsequent drying. The measured values in the lateral dimension are between 2 and 5 µN which is in the same range as expected by mechanical calculations. SEM micrographs of the damaged sites confirm a round breaking shape. The length of the broken line is around 0.5 to 1 µm with a lower limit of 0.5 µm which is similar to previously reported results on real damage produced by cleaning processes supported with Megasonic energy. The rupture in the AFM experiments occurs clearly in the polysilicon and not at the interfaces of the structure.
Enhanced particle removal processes in wet cleaning of semiconductor wafers can cause significant lateral forces on surface structures. These forces have to be limited to below the stability of the (nano) structures on the surface to prevent damage. A mechanical fracture test based on atomic force microscope manipulation was used in liquid media (deionized water, isopropanol, and propylene carbonate) to measure the lateral stability of polysilicon line structures. The results are compared to previously reported results in air. All liquid media investigated here showed a stabilizing effect. Maximum stability was found for immersion in isopropanol. The size of the damage generated was mainly influenced by the viscosity of the liquids. The results do not support a stress corrosion cracking process.
The fabrication of semiconductor devices includes the generation of high aspect ratio structures which are prone to lateral mechanical forces. Therefore, the mechanical stability of polysilicon line structures has been tested by an AFM technique for several line widths. Additionally, the damage has been evaluated and shows a uniform size distribution. These experimental results have been compared to numerical models, in which the influence of specific geometries and material present in the stack has been studied. A stress concentration region at the bottom of the line was observed and could be removed by corner rounding (fillet). Besides, the maximum stress is shifted away from the substrate. The experimental damage showed a stump at the bottom of the damage, thus confirming the numerical results.
Pattern collapse has long been known in photoresist patterning where the resist patterns merge or collapse during rinsing and drying steps [. The forces responsible for this collapse were identified as capillary forces during the drying process. Structures such as titanium nitride DRAM cylinders [ and silicon Flash shallow trench isolation (STI) lines have also been observed to be pattern collapse sensitive due to increase in aspect ratio of the features. Micro-electromechanical systems (MEMS) devices also show a similar phenomenon, but on a larger length scale, and is referred to as stiction [. For the technology nodes <14 nm, back-end-of-line (BEOL) low-k structures are also on the verge to show pattern collapse behavior. Whether a structure is sensitive to pattern collapse or not depends on several parameters, which will be analyzed in this paper.
Enhanced particle removal processes in wet cleaning as well as drying processes of semiconductor wafers can cause significant lateral forces on surface structures. These forces, however, must not exceed the mechanical stability of structures on the wafer. Thus, a mechanical fracture test was used to assess the lateral mechanical stability of polysilicon line structures in relevant process liquids. The mechanical test was based on nanomanipulation with an atomic force microscope. Compared to fracture tests in air, data acquired in liquid isopropanol revealed a stabilizing effect. The differences in fracture forces can be attributed to a stress corrosion process on the surface. The size of the generated damage was influenced by the viscosity of the surrounding media.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.