Quantitative {chi}{sup n} analysis of HREM images with applications to planar defects

Conference · Tue Dec 31 04:00:00 EST 1996

OSTI ID:468751

Zhang, H; Marks, L D ^[1]

---

1. Northwestern Univ., Evanston, IL (United States)

A number of different methods have been suggested in the literature for using HREM in a quantitative fashion, including R factors and cross-correlation analyses. The problem with many of these is that it is difficult to realistically gauge the errors involved when they are applied to real systems. Here we will discuss the approach based upon probabilistic {chi}{sup n} methods defined by: {chi}{sup n} = 1/m {Sigma} {vert_bar}I{sub c} - I{sub e}{vert_bar}{sup n}/{sigma}{sup n} where {sigma} is the error and m the number of degrees of freedom. The later must be independently determined for each picture, and the error in any measurement can then be strictly determined by a change in the calculated image such that {Delta}{chi}m = 1 (for 68% confidence level) and {chi}=1 for a good fit.

• OSTI ID: 468751
• Report Number(s): CONF-960877--
• Country of Publication: United States
• Language: English

Related Subjects

36 MATERIALS SCIENCE
40 CHEMISTRY
COPPER OXIDES
CRYSTAL DEFECTS
ELECTRON MICROSCOPY
ERRORS
HIGH-TC SUPERCONDUCTORS
IMAGES
MICROSTRUCTURE
STRONTIUM OXIDES