Auger electron angular distributions from surfaces: forward focusing or silhouettes?

Citation metadata

From: Science(Vol. 248, Issue 4959)
Publisher: American Association for the Advancement of Science
Document Type: Article
Length: 4,297 words

Document controls

Main content

Article Preview :

Auger Electron Angular Distributions from Surfaces: Forward Focusing or Silhouettes?

The recent article by Douglas G. Frank et al. [1] contains a number of conceptual errors that undermine both the data interpretation and the conclusions. The authors base their analysis on an erroneous set of notions about electron atom scattering. At the center of their misunderstanding is the statement that Auger emission events are uncorrelated and therefore cannot undergo the "formation of plane waves required for efficient diffraction." They then go on to imply that photoemission events are correlated and that the resulting photoelectrons can thereby undergo coherent diffraction. In point of fact, both Auger emission and photoemission events are uncorrelated. Also, correlation between Auger or photoemission events is not required to realize diffraction. Diffraction is nothing more than elastic scattering and interference, and all that is required for Auger or photoelectron diffraction to be detectable in an angle-resolved measurement is that the emitter be situated in a single crystal. In such a situation, the observed angular distributions show considerable intensity modulation as a result of the interference of the unscattered wave portion and all elastically scattered wave portions at the detector point.

Frank et al. then go on to claim that the modulation is caused entirely by "shadowing" (inelastic scattering). Yet, elastic scattering cross sections show considerable angular dependence and are largely peaked in the forward direction for all but the lowest kinetic energies. In contrast, there is no convincing evidence that inelastic scattering of low to medium kinetic energy electrons at single-crystal surfaces shows any anisotropy. The primary loss mechanism is plasma excitation, which is largely delocalized. A much weaker loss mechanism is excitation of bound core states by dipole scattering, which might be expected to show some angular dependence. In making their assertions, Frank et al. tacitly ignore 10 years of successful application of elastic scattering theory to the interpretation of angle-resolved Auger and photoelectron spectroscopic data. In rationalizing their results, they also ignore the well-established fact that Auger electrons and photoelectrons of the same kinetic energy from the same specimen exhibit nearly identical angular distributions, all of which are very well predicted by elastic scattering theory [2].

So why do the data of Frank et al. show minima along interatomic directions in Pt(111)? One possibility is that their specimen was not properly oriented about the surface normal during the measurements [3]. If their crystal was rotated 60[degrees] about the surface normal relative to where they thought it was, the low-energy electron diffraction pattern would not change, but the Auger intensity pattern would be inverted. If this error was made, what was interpreted to be electron intensity poking through the spaces between surface atoms would actually be forward-scattering--induced maxima along interatomic vectors. The latter interpretation is certainly much more consistent with basic principles of quantum-mechanical scattering than the proposition they forwarded.


[1] D. G. Frank et al., Science 247, 182 (1990).

[2] P. J. Orders, R. E. Connelly, N. F. T. Hall, C. S. Fadley, Phys. Rev. B...

Source Citation

Source Citation   

Gale Document Number: GALE|A9108827