1984
DOI: 10.1002/pssa.2210830134
|View full text |Cite
|
Sign up to set email alerts
|

The cathodoluminescence contrast formation of localized non-radiative defects in semiconductors

Abstract: The contrast formation of localized non-radiative semiconductor defects in cathodoluminescence (CL) micrographs is analyzed using a linear relationship between the CL intensity and the excess carrier density 8p. The latter is evaluated on the basis of Donolato's Born approximation and for generation by a point source. Numerical calculations for a threading dislocation a t right angle to a surface of infinite recombination velocity yield an exponential decay of the CL contrast a t sufficient distance from the d… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1

Citation Types

0
4
0

Year Published

1987
1987
2020
2020

Publication Types

Select...
6

Relationship

0
6

Authors

Journals

citations
Cited by 19 publications
(4 citation statements)
references
References 4 publications
0
4
0
Order By: Relevance
“…Since the exact dislocation carrier capture time is not known, we assume here a capture time of the order of a few picoseconds, comparable with the one into QD states [38], [39]. In reality, the physical radius of an actual dislocation core is estimated to be only a few tens of nanometers [40], [41], so varying dis is not only a means of modelling realistic performance trends with growing dislocation density, but also a way to compensate for larger z, which overestimate the dislocation size slightly. For z = 500 nm, a value of dis = 10 ps is chosen to model dislocation sections, whereas dis is set to infinity in dislocation-free regions.…”
Section: Implementation Of Dislocationsmentioning
confidence: 99%
“…Since the exact dislocation carrier capture time is not known, we assume here a capture time of the order of a few picoseconds, comparable with the one into QD states [38], [39]. In reality, the physical radius of an actual dislocation core is estimated to be only a few tens of nanometers [40], [41], so varying dis is not only a means of modelling realistic performance trends with growing dislocation density, but also a way to compensate for larger z, which overestimate the dislocation size slightly. For z = 500 nm, a value of dis = 10 ps is chosen to model dislocation sections, whereas dis is set to infinity in dislocation-free regions.…”
Section: Implementation Of Dislocationsmentioning
confidence: 99%
“…The dislocations may have their own energy levels in the band gap and. respectively, specific lines in the CL spectrum (Lohnert andKubalek 1984, Yacobi andHold 1986). In most cases, the dislocations are nonluniinescent.…”
Section: Introductionmentioning
confidence: 99%
“…This law applies to the one-dimensional case, whereas in three dimensions p actually decreases as (l/r) exp (-r/L); as a consequence of the geometrical factor l/r, the extent of the cloud of beam-injected minority carriersand hence the resolution of the CL imageremains of the order of Re, even in the limit of L + co [2]. This point has been elucidated in connection with the modeling of the EBIC images of semiconductor defects [2, 31, and is mentioned in related review articles [9 to 111, but apparently its relevance for CL imaging has been overlooked, in spite of the similarity between the contrast formation of the two techniques [12].In this connection, it should be noted that in [l] the diffusion equation (1) as it stands is clearly incorrect; moreover, the claim that this equation has been analytically solved only for the point source or the uniform generation sphere seems to neglect, for instance, the solution given in [13] for generation with a Gaussian lateral distribution having depth -dependent amplitude and variance.…”
mentioning
confidence: 99%
“…This law applies to the one-dimensional case, whereas in three dimensions p actually decreases as (l/r) exp (-r/L); as a consequence of the geometrical factor l/r, the extent of the cloud of beam-injected minority carriersand hence the resolution of the CL imageremains of the order of Re, even in the limit of L + co [2]. This point has been elucidated in connection with the modeling of the EBIC images of semiconductor defects [2, 31, and is mentioned in related review articles [9 to 111, but apparently its relevance for CL imaging has been overlooked, in spite of the similarity between the contrast formation of the two techniques [12].…”
mentioning
confidence: 99%