Answers are grounded in this paper and cite their sources.
The total fraction of the current transmitted and its angular distribution have been measured when an electron beam is incident normally on evaporated films of aluminium, copper, silver and gold. The range of mass-thickness ρ x investigated was such that plural scattering prevailed in the thinnest, and diffusion conditions in the thickest films. The energy of the incident beam was varied between 5 and 30 keV. The experimental results are not explained in quantitative detail by any of the existing theories. Multiple scattering, as defined by the angular distribution being Gaussian, sets in at a mass-thickness corresponding to 25 ± 5 elastic scattering acts for all four elements, at both 10 and 20 keV. The peak of the angular distribution occurs at an angle λ A which is one-half to one-third that predicted by Bothe (λ B ) and still appreciably smaller than that given by the Molière-Lenz corrections. The total fraction η T transmitted through the thinner films agrees, to first order, equally well with the rival multiple scattering theories of Bothe and of Bethe, Rose and Smith. The parameter λ X , which satisfies Bothe's equation for η T , is generally found to be well below his value λ B and does not vary in the predicted way with E 0 and Z . Thus, in this energy range, λ A and λ X are no longer identical as they are in the theory. The probable reason is that an appreciable fraction of the beam is lost by backscattering even from very thin films, i.e. large angle scattering is more important than Bothe allowed. However, the fact that both λ A 2 and λ X 2 are proportional to ρ x indicates that Bothe's statistical treatment of multiple scattering is basically correct. The Molière version of it is inadequate, being applicable only over a narrow range of mass-thickness at these low energies; also it appears to give over-correction both with increasing ρ x and for the Born approximation. In thicker films, where diffusion prevails, the angular distribution is closer to the cos 2 θ form of Bothe's diffusion theory than to the cos θ form predicted by Bethe, Rose and Smith. λ A tends to a limiting value slightly larger than that given by the Bothe theory. The fraction transmitted begins to obey an exponential law at a mass-thickness which varies with E 0 and Z . This critical thickness, expressed as a fraction of the range, increases much more rapidly for Au and Ag than for Cu and Al as E 0 is reduced; at 5 keV an exponential fall occurs only towards the end of the range for all four elements. The values of the absorption coefficient μ X are much smaller (one-half to one-fifth) than Bothe's diffusion theory predicts, and increase at differing rates for each element with falling E 0 , until at 5 keV all four elements have the same value. The depth x D at which diffusion sets in has been determined on the basis of four different definitions. The values for each element agree among themselves within experimental error, and are in closer agreement with Tomlin's than with Archard's model of electron penetration. The empirical relations established for the measured quantities, λ A , λ X , μ X and x D , show a much less marked variation with both E 0 and Z than in Bothe's theories; Molière's treatment reduces the dependence on Z , but not sufficiently. It is concluded that the Wentzel differential scattering cross section, on which these theories rest, is inadequate at low E 0 and needs to be replaced by a formulation which accounts better for large-angle as well as small-angle scattering.
No takes yet. Share an insight, caveat, or question.
Cosslett et al. (1964) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: