Convergent beam electron diffraction ͑CBED͒ is used in this study to investigate the stress distribution around shallow trench isolation ͑STI͒ structures. Attention is given to the influence of the different processing parameters and the width and spacing of the structures. The use of a wet or a dry pregate oxidation is found to have a strong influence on the stress behavior. Isolated lines show more stress, leading to the formation of defects in the silicon substrate if a wet pregate oxidation is used. The CBED analyses are compared with micro-Raman and bright-field transmission electron microscopy measurements.Shallow trench isolation ͑STI͒ is most frequently used as a lateral isolation module in deep submicrometer technologies. 1 The various processing steps cause stress fields in the STI structures, which can lead to defect formation in the silicon substrate. In their turn, stress fields affect the electrical parameters and the reliability of devices. The stress depends in addition to the processing parameters, in a large measure on the width and spacing of the STI structures.Micro-Raman spectroscopy has been a routine tool for stress evaluation of microelectronic structures 2 but fails for the currently fabricated, highly integrated 0.18 m ULSI ͑ultralarge scale integration͒ devices because of its spatial resolution, which is limited by its probe size to about 1 m. Convergent beam electron diffraction ͑CBED͒, on the contrary, has a very high spatial resolution ͑limited by its probe size and the necessary specimen tilt͒. In this study, a resolution of 30 nm is obtained. CBED is shown to be a powerful technique to investigate local stress fields in small devices in a quantitative way. [3][4][5] CBED uses a convergent incident electron beam, resulting in a diffraction pattern consisting of disks rather than spots, as is the case for diffraction with a parallel incident beam. The central disk of the CBED pattern contains a fine structure: deficiency lines corresponding with reflections at higher order Laue zones ͑HOLZ lines͒. These lines are very sensitive to lattice parameter and electron-beam voltage variations and can therefore be used for stress evaluation.In this study, stress analyses with CBED are performed around nonoptimized STI structures with various processing conditions and linewidths. ExperimentalSTI processing.-The STI structures 1 are processed on 200 mm ͑100͒-oriented p-type silicon substrates. After the deposition of a 150 nm low-pressure chemical vapor deposited ͑LPCVD͒ nitride film on a 15 nm thermal pad oxide, the active area is defined by a standard photolithography step. Subsequently, the nitride/oxide stack is etched and 0.4 m deep trenches are etched in the silicon substrate. After cleaning, a cavity etch in the pad oxide and a sidewall oxidation of 40 nm are performed. 6 The trenches are refilled using a 450 nm high-density plasma ͑HDP͒ oxide and a 100 nm LPCVD nitride. This second nitride layer is again defined by a photolithographic step as a field protecting nitride and is etched sel...

No abstract

State-of-the-art semiconductor devices require an accurate control of the complete twodimensional dopant distribution. The routine use of process simulators to predict the envisaged distributions and their resulting accuracy, is strongly linked to the physical models contained in these programs as well as their calibration. Whereas SIMS and SRP have been used extensively for the calibration of 1D-profiles, calibration of 2D-profiles has been very limited.In this work, we report some results obtained with the 2D-profiling techniques SSRM (Scanning Spreading Resistance Microscopy) and SCM (Scanning Capacitance Microscopy ) for the study of two-dimensional effects on diffusion. In particular, we discuss the role of the nitride spacer on the lateral diffusion of arsenic and boron. Using series of transistors with different nitride spacers with or without TEOS-liners, a strong dependence between the lateral diffusion and the nitride spacer thickness can be observed using SSRM and SCM. The process flow eliminates the possible contribution of Transient enhanced diffusion (TED) as a dominant mechanism. At the same time an enhancement of the lateral stress underneath the spacers has been observed with CBED and Raman, suggesting a correlation between the lateral diffusion and the nitride spacers. The enhanced diffusion of As and B is strongly linked to the spacer size whereby the differences in enhancement suggest that the proximity of the dopants to the stress field field region is an important parameter.

Initial studies (using Scanning Spreading Resistance Microscopy) on the lateral diffusion of B and As have shown an important influence of the thickness of oxy/nitride spacers. The latter phenomenon was tentatively ascribed to stress enhanced diffusion under the spacer region [1]. These studies have been complemented with Scanning Capacitance Microscopy (SCM) measurements, which confirm the SSRM-data. In fact both techniques shows a similar increase in lateral diffusion with increasing spacer thickness (∼ 0.2 nm/nm spacer thickness), whereby no effect is observed on the vertical diffusion. When using spacers with or without TEOS-liner, fairly similar enhancements could be seen. Micro-Raman and CBED stress measurements for these cases do however show a large reduction in stress when a TEOS-liner is used, suggesting that the correlation (at least to the final) stress is not really justified. A possible explanation could however be that the lateral diffusion occurs before the stress relaxation within the thermal treatment. In order to elucidate the diffusion mechanism (initial stress, interstitials, hydrogen incorporation, TED,..) we have expandedthe experimental matrix with a vacancy diffuser such as Sb and simulated the potential H-incorporation duringthe nitride deposition by a hydrogen anneal. Moreover we also have studied the impact of TED by splits with RTP-anneals before the nitride deposition.

No abstract