Shallow trench isolation ͑STI͒ allows tighter device packing and reduced chip area for isolation. STI is critically dependent on the global planarity that is only possible using chemical mechanical polishing ͑CMP͒. Ceria-based slurries are considered the most promising candidates for STI CMP. Despite decades of use in glass polishing, the unique characteristics of ceria slurries are not well understood. In this study, we have conducted force measurements and tribological tests using an atomic force microscope ͑AFM͒ and a scanning electron microscope to investigate pH-dependent ceria-silica and silica-silica interactions that occur during CMP. Our studies confirm the effect of hydrolysis at high pH during silica-silica abrasion. An additional physicochemical contribution to ceria-silica polishing is identified and discussed. Furthermore, a strong correlation was observed between the AFM based studies and in situ friction force measurements during CMP.The increased packing density and ever shrinking dimensions of modern integrated circuits are direct consequences of what is commonly termed Moore's Law. 1 Shallow trench isolation ͑STI͒ is a novel manufacturing technique that facilitates the miniaturization of devices. In the STI process, trenches are etched into the Si 3 N 4 mask layer and the silicon substrate. Chemical vapor deposition ͑CVD͒ or spin-on-glass ͑SOG͒ techniques are used to fill the trenches with SiO 2 . CMP or a combination of reactive ion etching ͑RIE͒ and chemical mechanical planarization ͑CMP͒ then remove the overlying oxide. The planarized, trench-fill oxide isolates the device after removal of the Si 3 N 4 mask.STI is not possible without CMP. An ideal STI CMP process stops at the nitride layer, resulting in uniform removal of the trenchfill oxide. Naturally, STI CMP slurries must preferentially polish oxide with respect to the nitride layer, and should leave the surface globally planar and defect free. Over-polishing of the nitride or erosion of the oxide causes damage to the active region and degrades isolation performance. 2 Traditional interlayer dielectric consumables are not sufficient for the STI CMP process. 3 New slurry and pad chemistries with tunable properties are required to meet the stringent selectivity, planarity, removal rate, and surface finish requirements of the STI process. Perhaps the most important requirement is a high oxide to nitride polish rate. For STI CMP, ceria-based slurries outperform silica-based slurries. 4 Ceria slurries preferentially remove the oxide layer with minimal polishing of the underlying nitride. Despite the use of ceria abrasives in glass polishing for many decades, the ceria-silica polishing mechanism for STI CMP remains unclear.Ceria-based slurries are capable of silica removal rates ϳ5ϫ that of conventional silica slurries. This enhancement is even more pronounced when the silica removal rate is normalized to the respective solids loading ͑ϳ2% for CeO 2 slurries and ϳ10% for SiO 2 slurries͒. Cook attributed the enhanced polishing of glass by ceri...

Chemical mechanical polishing (CMP) is widely adopted in producing excellent local and global planarization of microelectronic devices. However, the fundamental mechanisms of material removal and the interactions of the chemical and mechanical effects are not well understood. In the present paper, the contact area of a pad with a wafer is measured in dry and wet conditions in different pH solutions using optical microscopy and Fourier transform infrared spectroscopy, respectively. Pad surface mechanical properties in dry and wet states are also investigated using atomic force microscopy. The results indicate a significant difference in pad surface mechanical properties between dry and wet states, which could be correlated with the CMP removal rates.

For highly selective particle-based slurries or fixed abrasive pads, ceria has been identified as the abrasive of choice for the chemical mechanical polishing (CMP) step for shallow trench isolation (STI). The advantageous performance of ceria-based CMP consumables is usually attributed to enhanced chemical reactivity between ceria and oxide materials. In fact, this reaction is a central theme of all ceria polishing models from glass polishing to STI CMP. Previously, experimental evidence in support of the ceria-silica reaction during CMP was virtually non-existent. Recently, we proposed a pH-dependent ceria-silica polishing mechanism based on polishing results, in-situ friction force measurements, and spectroscopic and microscopic investigations. In this report, we have studied the chemical interactions between ceria and silica in the absence of particles using an atomic force microscope (AFM) and a scanning electron microscope (SEM). AFM silicon tapping mode cantilevers were functionalized by depositing a silica coating via chemical vapor deposition (CVD) and thermal oxidation. SEM imaging and compositional analysis was conducted on the cantilevers before and after wear against a ceria thin film, which was grown by pulsed laser deposition. The cantilever wear profile and elemental composition as a function of pH confirms our earlier polishing results and the pH-dependent CMP model for ceria-silica CMP.

No abstract

Chemical mechanical polishing (CMP) is widely used for local and global planarization of microelectronic devices. It has been demonstrated experimentally in the literature that the polishing performance is a result of the synergistic effect of both the chemicals and the particles involved in CMP. However, the fundamental mechanisms of material removal and the interactions of the chemical and mechanical effects are not well understood. A comprehensive model for CMP was developed taking into account both the chemical and mechanical effects for monodisperse slurries. The chemical aspect is attributed to the chemical modification of the surface layer due to slurry chemistry, whereas the mechanical aspect is introduced by indentation of particles into the modified layer and the substrate depending on the operating conditions. In this study, the model is extended to include the particle size and pad asperity distribution effects. The refined model not only predicts the overall removal rate but also the surface roughness of the polished wafer, which is an important factor in CMP. The predictions of the model show reasonable agreement with the experimental observations.