scholarly journals HRTEM image simulations of structural defects in gate oxides

Author(s):  
S. Taylor
2000 ◽  
Vol 6 (S2) ◽  
pp. 1078-1079
Author(s):  
S. Taylor ◽  
J. Mardinly ◽  
M.A. O'Keefe ◽  
R. Gronsky

High-resolution transmission electron microscopy (HRTEM) is used extensively in the semiconductor industry for device characterization, and has become one of the highly favored techniques for characterizing the latest generation of ultra-thin gate oxides in MOSFET devices. However, relatively little is understood (either quantitatively or experimentally) about the limitations of HRTEM in detecting structural defects in gate oxides that could affect device performance. To investigate model defects experimentally, it would be necessary to construct “perfect” gate oxides, introduce defects with size and morphology known perfectly a priori, successfully make thin specimens that capture the defects, and then perform imaging experiments in the HRTEM. Since that task is virtually impossible, we have performed HRTEM image simulations to assess the visibility of various structural defects in gate oxides. The gate oxide was modeled as an amorphous silicon oxide 16.3Å-thick, sandwiched between a gate and substrate. The substrate was (100) silicon viewed along the [110] direction.


2002 ◽  
Vol 8 (5) ◽  
pp. 412-421 ◽  
Author(s):  
Seth T. Taylor ◽  
John Mardinly ◽  
Michael A. O'Keefe

We have performed high resolution transmission electron microscope (HRTEM) image simulations to qualitatively assess the visibility of various structural defects in ultrathin gate oxides of MOSFET devices, and to quantitatively examine the accuracy of HRTEM in performing gate oxide metrology. Structural models contained crystalline defects embedded in an amorphous 16-Å-thick gate oxide. Simulated images were calculated for structures viewed in cross section. Defect visibility was assessed as a function of specimen thickness and defect morphology, composition, size, and orientation. Defect morphologies included asperities lying on the substrate surface, as well as “bridging” defects connecting the substrate to the gate electrode. Measurements of gate oxide thickness extracted from simulated images were compared to actual dimensions in the model structure to assess TEM accuracy for metrology. The effects of specimen tilt, specimen thickness, objective lens defocus, and coefficient of spherical aberration (Cs) on measurement accuracy were explored for nominal 10-Å gate oxide thickness. Results from this work suggest that accurate metrology of ultrathin gate oxides (i.e., limited to several percent error) is feasible on a consistent basis only by using a Cs-corrected microscope. However, fundamental limitations remain for characterizing defects in gate oxides using HRTEM, even with the new generation of Cs-corrected microscopes.


2002 ◽  
Vol 17 (1) ◽  
pp. 204-213 ◽  
Author(s):  
G. Y. Yang ◽  
J. M. Finder ◽  
J. Wang ◽  
Z. L. Wang ◽  
Z. Yu ◽  
...  

Microstructure in the SrTiO3/Si system has been studied using high-resolution transmission electron microscopy and image simulations. SrTiO3 grows heteroepitaxially on Si with the orientation relationship given by (001)STO//(001)Si and [100]STO//[110]Si. The lattice misfit between the SrTiO3 thin films and the Si substrate is accommodated by the presence of interfacial dislocations at the Si substrate side. The interface most likely consists of Si bonded to O in SrTiO3. The alternative presentation of Sr and Si atoms along the interface leads to the formation of 2× and 3× Sr configurations. Structural defects in the SrTiO3 thin film mainly consist of tilted domains and dislocations.


2000 ◽  
Vol 6 (S2) ◽  
pp. 1080-1081
Author(s):  
S. Taylor ◽  
J. Mardinly ◽  
M.A. O'Keefe ◽  
R. Gronsky

High resolution transmission electron microscopy (HRTEM) has found extensive use in the semiconductor industry for performing device metrology and characterization. However, shrinking device dimensions (gate oxides are rapidly approaching 10Å) present challenges to the use of HRTEM for many applications, including gate oxide metrology. In this study, we performed HRTEM image simulations of a MOSFET device to examine the accuracy of HRTEM in measuring gate oxide thickness. Length measurements extracted from simulated images were compared to actual dimensions in the model structure to assess TEM accuracy. The effects of specimen tilt, specimen thickness, objective lens defocus and coefficient of spherical aberration (CS) on measurement accuracy were explored for nominal 10Å and 16Å gate oxide thicknesses.The gate oxide was modeled as an amorphous silicon oxide situated between a gate electrode and substrate, both modeled as single crystal Si(100). Image simulations of the sandwich structure were performed in cross-section (with Si[110] parallel to beam direction) using the multislice approximation for a 200 kV microscope with Cs=0.5mm.


1998 ◽  
Vol 4 (S2) ◽  
pp. 596-597
Author(s):  
D.L. Medlin ◽  
J.E. Smugeresky ◽  
D. Cohen

HRTEM images of α-Al2O3 oriented along [1100] often exhibit a modulation of the basal fringe intensity with a period that corresponds to twice the basal plane spacing. In the example shown in figure 1, note the strong basal fringe contrast in the A12O3 with a period of 4.33 Å. HRTEM image simulations for ideal imaging conditions fail to reproduce this doubling of the lattice fringe period. Instead, simulated images predict that the (0006) fringes (d0006=2.17 Å) should be of equal intensity (e.g., see Figure 3a). Incorrect beam and/or crystal tilt can strongly affect HRTEM image contrast. In particular, anomalous periodicities can arise under beam and crystal tilt conditions for which normally kinematically and dynamically forbidden diffracted beams contribute to the image. More recently, the importance of 3-fold astigmatism on HRTEM imaging has been recognized. As with beam tilt, this lens aberration produces asymmetrical phase shifts in the diffracted beams.


Author(s):  
Jan-Olle Malm ◽  
Michael A. O’Keefe

In all HRTEM images, the incident electron beam suffers from the effects of limited spatial and temporal coherence. These effects produce a smearing of the image, and provide the ultimate limits as to how high a spatial frequency can be transferred to the image (i.e. resolution). The effect of partial temporal coherence is manifested as a spread of focus, and that of partial spatial coherence as incident beam convergence. The effects of partial coherence can be included in HRTEM image simulations by summing images in real space, or by applying an appropriate transmission cross-coefficient (TCC) when computing the image intensity spectrum in reciprocal space. Figure 1 shows a method of including the effects of incident beam convergence by real-space summation. The degree of convergence is estimated by measuring the spot size in the experimental diffraction pattern (a). Each spot in the diffraction pattern is sub-sampled (b), and a series of images is computed at incident beam angles appropriately sampling the convergence cone.


Author(s):  
J. Kioseoglou ◽  
G. P. Dimitrakopulos ◽  
Th. Kehagias ◽  
E. Kalessaki ◽  
Ph. Komninou ◽  
...  

2003 ◽  
Vol 9 (S03) ◽  
pp. 204-205
Author(s):  
Johannes Biskupek ◽  
Ute Kaiser ◽  
Andrey Chuvilin

Sign in / Sign up

Export Citation Format

Share Document