Secondary ion mass spectrometry (SIMS) is still one of the reference analytical techniques in order to measure depth distribution of impurities in silicon due to its extreme sensitivity, excellent depth resolution and good reproducibility. However, when facing ultra shallow junctions for nowadays source and drain extension in Si CMOS devices, several issues need to be addressed. In the case of boron depth profiling, a detailed literature has reported oxidizing conditions like normal O2+ ion bombardment or oblique incidence with 'oxygen flooding' (a controlled pressure/leak of oxygen in the analysis ambient) as effective in terms of detection limit, dosimetry accuracy and depth resolution despite an often observed shift of B profiles toward the surface. However, it is known that the profile shape at the surface and/or at the native SiO2/ Si interface is strongly influenced by artefacts resulting in a B peak at the surface. In recent years, a not-oxidizing approach resulted more accurate in revealing the profile shape as revealed by comparing SIMS results with nuclear techniques such as ERDA. The drawback of this approach is represented by the rapid ripple and roughness formation on the SIMS crater bottom for those sputtering conditions (under-keV impact energy and oblique incidence): the effect is a rapid variation of sputtering yield hindering the accuracy of the depth calibration if a constant sputter rate (SR) is applied [4]. In this work we implemented in the not-oxidizing approach the use of Zalar rotation, i. e. the sample was kept rotating during the oblique incidence O2+ sputtering in order to reduce the roughness formation and prevent the SR variations. Samples were 11B ultra low energy implants in single crystal silicon with implant energy varying between 0.2 and 3.0 keV and implanted fluency between 1E14 and 5E15 cm-2. Analysis was carried out using either a 0.5 or 0.3 keV impact energy O2+ beam with ~70° incidence and collecting positive secondary ions. The resulting quantified profiles were crosschecked with results of soft x-ray grazing incidence x-ray fluorescence (SR-GIXRF) obtained at Bessy II synchrotron facility in Berlin within EC financed ANNA project (contract n. 026134(RII3)). Different quantification issues will be reported and discussed.
Ultra low energy Boron implants in silicon characterization by not-oxidizing secondary ion mass spectrometry analysis and soft-ray grazing incidence x-ray fluorescence techniques.
Giubertoni, Damiano;Pepponi, Giancarlo;Iacob, Erica;Gennaro, Salvatore;Bersani, Massimo
2009-01-01
Abstract
Secondary ion mass spectrometry (SIMS) is still one of the reference analytical techniques in order to measure depth distribution of impurities in silicon due to its extreme sensitivity, excellent depth resolution and good reproducibility. However, when facing ultra shallow junctions for nowadays source and drain extension in Si CMOS devices, several issues need to be addressed. In the case of boron depth profiling, a detailed literature has reported oxidizing conditions like normal O2+ ion bombardment or oblique incidence with 'oxygen flooding' (a controlled pressure/leak of oxygen in the analysis ambient) as effective in terms of detection limit, dosimetry accuracy and depth resolution despite an often observed shift of B profiles toward the surface. However, it is known that the profile shape at the surface and/or at the native SiO2/ Si interface is strongly influenced by artefacts resulting in a B peak at the surface. In recent years, a not-oxidizing approach resulted more accurate in revealing the profile shape as revealed by comparing SIMS results with nuclear techniques such as ERDA. The drawback of this approach is represented by the rapid ripple and roughness formation on the SIMS crater bottom for those sputtering conditions (under-keV impact energy and oblique incidence): the effect is a rapid variation of sputtering yield hindering the accuracy of the depth calibration if a constant sputter rate (SR) is applied [4]. In this work we implemented in the not-oxidizing approach the use of Zalar rotation, i. e. the sample was kept rotating during the oblique incidence O2+ sputtering in order to reduce the roughness formation and prevent the SR variations. Samples were 11B ultra low energy implants in single crystal silicon with implant energy varying between 0.2 and 3.0 keV and implanted fluency between 1E14 and 5E15 cm-2. Analysis was carried out using either a 0.5 or 0.3 keV impact energy O2+ beam with ~70° incidence and collecting positive secondary ions. The resulting quantified profiles were crosschecked with results of soft x-ray grazing incidence x-ray fluorescence (SR-GIXRF) obtained at Bessy II synchrotron facility in Berlin within EC financed ANNA project (contract n. 026134(RII3)). Different quantification issues will be reported and discussed.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.