Atom Probe Tomography Analysis of Thick Film SiO 2 and Oxide …€¦ · In this presentation, we...
2
Atom Probe Tomography Analysis of Thick Film SiO 2 and Oxide Interfaces: Conditions Leading to Improved Analysis Yield T.J. Prosa, D. Lawrence, D. Olson, D. Lenz, J.H. Bunton, and D.J. Larson Cameca Instruments Inc., 5500 Nobel Drive, Madison, WI, 53711 Atom Probe Tomography (APT) is a time-of-flight mass spectrometry imaging technique that establishes extreme surface electric fields (~20 V/nm) on a specimen to initiate field-evaporation of surface material [1]. Consequently, it is somewhat surprising that a number of recent results have established that large band-gap dielectric materials (electrical insulators) can be analyzed by laser- pulsed APT, especially when the bulk band-gap of the material readily exceeds the photon energy of the laser-pulse used to heat the sample. These dielectric materials include but are not limited to Al 2 O 3, SiO 2 , ZnO, MgO, and olivine [2-6]. The analysis of native and engineered silicon oxide films of varying thickness is of interest because these structures play a critical role in device performance for current generation microelectronics [7]. Unfortunately, such structures often result in APT specimens that are prone to failure under typical analysis conditions which has undoubtedly slowed the adoption of this technique within the microelectronics industry. It is therefore essential that APT measurements are optimized, not only to ensure controlled field-evaporation resulting in data with adequate spatial and chemical resolution to impact device development, but also to maximize the potential for specimen survivability (analysis yield) to facilitate increased utility of the technique. In this presentation, we report on our efforts to optimize APT analysis of thick film SiO 2 and Si/SiO 2 interfaces. A test structure made from a 120 nm thick SiO 2 film thermally grown on <100> silicon and subsequently capped with 300 nm of 650 o C poly-Si has been used to investigate the tradeoffs between various analysis conditions (wavelength, spot size, polarization, laser-pulse energy, evaporation rate, voltage/field control algorithm, etc.) and their effects on data quality and analysis yield. For top-down , changes in specimen temperature are monitored by variations in the Si charge-state-ratio (CSR). The most pronounced variation in CSR occurs during analysis of the poly-Si layer. Evolution of tip size and shape creates a quantum-dot- like volume of material isolated above the 120 nm thick oxide, which likely affects the photon absorption as the region evolves in size and shape. Under appropriate analysis conditions, high quality mass spectra (Si ++ mass resolving power FWHM of ~1200) can be achieved for the LEAP ® 4000X HR (see Fig. 1). In addition, we have achieved good analysis yield (>33%) when per-pulse evaporation rates are carefully controlled through the Si/SiO 2 interface regions. References [1] M.K. Miller et al., Atom Probe Field Ion Microscopy, Oxford University Press, Oxford, 1996. [2] E.A. Marquis et al., Materials Today 13 (2010) 34. [3] D.J. Larson et al., Microsc. Microanal. 14 (2008) 1254. [4] E. Talbot et al., EPL 87 (2009) 26004. [5] Y.M. Chen et al., Ultramicroscopy (2010). [6] S.W. Parman et al., Fall Meeting, AGU, San Francisco, CA, Abstract V43F-07 (2010). [7] M.T. Bohr et al., IEEE Spectrum 44 (2007) 29. 750 doi:10.1017/S1431927611004624 Microsc. Microanal. 17 (Suppl 2), 2011 © Microscopy Society of America 2011 https://doi.org/10.1017/S1431927611004624 Downloaded from https://www.cambridge.org/core. IP address: 54.39.106.173, on 16 Jan 2021 at 11:12:59, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
Transcript of Atom Probe Tomography Analysis of Thick Film SiO 2 and Oxide …€¦ · In this presentation, we...
Atom Probe Tomography Analysis of Thick Film SiO2 and Oxide Interfaces: Conditions Leading to Improved Analysis Yield T.J. Prosa, D. Lawrence, D. Olson, D. Lenz, J.H. Bunton, and D.J. Larson Cameca Instruments Inc., 5500 Nobel Drive, Madison, WI, 53711 Atom Probe Tomography (APT) is a time-of-flight mass spectrometry imaging technique that establishes extreme surface electric fields (~20 V/nm) on a specimen to initiate field-evaporation of surface material [1]. Consequently, it is somewhat surprising that a number of recent results have established that large band-gap dielectric materials (electrical insulators) can be analyzed by laser-pulsed APT, especially when the bulk band-gap of the material readily exceeds the photon energy of the laser-pulse used to heat the sample. These dielectric materials include but are not limited to Al2O3, SiO2, ZnO, MgO, and olivine [2-6]. The analysis of native and engineered silicon oxide films of varying thickness is of interest because these structures play a critical role in device performance for current generation microelectronics [7]. Unfortunately, such structures often result in APT specimens that are prone to failure under typical analysis conditions which has undoubtedly slowed the adoption of this technique within the microelectronics industry. It is therefore essential that APT measurements are optimized, not only to ensure controlled field-evaporation resulting in data with adequate spatial and chemical resolution to impact device development, but also to maximize the potential for specimen survivability (analysis yield) to facilitate increased utility of the technique. In this presentation, we report on our efforts to optimize APT analysis of thick film SiO2 and Si/SiO2 interfaces. A test structure made from a 120 nm thick SiO2 film thermally grown on <100> silicon and subsequently capped with 300 nm of 650 oC poly-Si has been used to investigate the tradeoffs between various analysis conditions (wavelength, spot size, polarization, laser-pulse energy, evaporation rate, voltage/field control algorithm, etc.) and their effects on data quality and analysis yield. For top-down , changes in specimen temperature are monitored by variations in the Si charge-state-ratio (CSR). The most pronounced variation in CSR occurs during analysis of the poly-Si layer. Evolution of tip size and shape creates a quantum-dot-like volume of material isolated above the 120 nm thick oxide, which likely affects the photon absorption as the region evolves in size and shape. Under appropriate analysis conditions, high quality mass spectra (Si++ mass resolving power FWHM of ~1200) can be achieved for the LEAP® 4000X HR (see Fig. 1). In addition, we have achieved good analysis yield (>33%) when per-pulse evaporation rates are carefully controlled through the Si/SiO2 interface regions. References [1] M.K. Miller et al., Atom Probe Field Ion Microscopy, Oxford University Press, Oxford, 1996. [2] E.A. Marquis et al., Materials Today 13 (2010) 34. [3] D.J. Larson et al., Microsc. Microanal. 14 (2008) 1254. [4] E. Talbot et al., EPL 87 (2009) 26004. [5] Y.M. Chen et al., Ultramicroscopy (2010). [6] S.W. Parman et al., Fall Meeting, AGU, San Francisco, CA, Abstract V43F-07 (2010). [7] M.T. Bohr et al., IEEE Spectrum 44 (2007) 29.
750doi:10.1017/S1431927611004624
Microsc. Microanal. 17 (Suppl 2), 2011© Microscopy Society of America 2011
https://doi.org/10.1017/S1431927611004624Downloaded from https://www.cambridge.org/core. IP address: 54.39.106.173, on 16 Jan 2021 at 11:12:59, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
FIG. 1. Mass spectra and atom map (inset) for each region of the 120 nm thick oxide containing structure. Ions are color coded as shown in each mass spectrum: a) The crystal Si-substrate region exhibiting Si ions in the 1+ and 2+ charge states, b) the poly-Si region containing a small number of oxygen ions, and c) the SiO2 region exhibiting Si, O, O2, and other Si-O complex ions.
1e5
1e4
1e3
1e2
Cou
nts
Mass-to-Charge Ratio (Da)10 20 30 40 50 60
Si++
Si+
Si+++
SiO++SiO+
O+
SiO2+
O++
1246 FWHM651 FW0.1M122 FW0.01M
1245 FWHM600 FW0.1M117 FW0.01M
O2+
Si++1e6
1e4
1e2
15 25 35 45
Si+
1281 FWHM683 FW0.1M399 FW0.01M
a)
1e5
1e3
1e1
15 25 35 45
Si++ Si+
1147 FWHM580 FW0.1M302 FW0.01M
b)
c)
180 160 140 120 100 80 60 40 20Z (nm)
SiO2
poly-Si
Si substrate
Microsc. Microanal. 17 (Suppl 2), 2011 751
https://doi.org/10.1017/S1431927611004624Downloaded from https://www.cambridge.org/core. IP address: 54.39.106.173, on 16 Jan 2021 at 11:12:59, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
