Backscattered electrons properties
Some results from "Monte Carlo simulation of electron transport", F. Salvat, J. M. Fernandez-Varea and R. Mayol, in "Electron Microscopy in Materials Science", editors P. G. Merli and M. Vittori Antisari, published by World Scientific (1991).
The above image shows the energy distribution of the BS electrons for 2 beam incidence angles. The calculated values (in the histogram) are compared with the experimentally measured values.
The above images shows the energy and angular distributions of transmitted electrons through 3 diferent thin foils of Al. Experimental results (in the form of continuous curves) are shown for comparison. These are similar to the information obtained from the backscattered electrons.
One of the main and first parameters used for testing Monte Carlo simulations is the backscattering coefficient, evaluated for different thick specimens (single materials) at different beam tilt angles. The following shows calculated results compared to one source of experimental data.
The following link contains a large amount of experimental data about BS and SE coefficients for many materials Experimental BS and SE data for many materials. This is used for comparison with calculated data, in the process of testing Monte Carlo calculation algorithms.
BS linescans across boundaries
The following data shows the effect of material parameters on the type of contrast obtained when scanning across a boundary between two materials with different atomic numbers (and other parameters). In this type of data the position of the beam is varied across the sample, and a plot is made of the intensity of a signal at all the calculated plots.
Results taken from: "A Monte Carlo Study of the Position of Phase Boundaries in Backscattered Electron Images", D. R. Cousens and D. C. Joy, in Scanning Vol 19, 547-552 (1997, in which a plural scattering simulation model was used.
The following linescan shows the effect of a finite beam size on BS electron contrast measurements (in this case the estimated beam size is about 60A FWHM, equivalent to a tungsten SEM filament). Data from: "BSE Image Simulation in Scanning Electron Microscopy" by Z. J. Radzimski, in Journal of Computer-Assisted Microscopy, Vol. 6, No. 4, 1994.
From the source the following image shows the BSE linescans for two very different structures (as used in semiconductor microfabrication). But, the corresponding BSE signals for them are very similar, for a range of beam conditions.
E-Beam lithography
In electron beam lithography it is important to know precisely the amount of energy deposited in the resist, as this will decide the dimensions of the developed features. Among the factors deciding the total exposure of the resist are the electrons scattered from the substrate layer, and exposure from adjacent regions (known generally as proximity effect).
For both problems above Monte Carlo simulation provides a practical answer. In some cases, to directly control the micro-fabricator, and in others, to calculate empyrical relations (such as the point spread function) to allow exposure correction under certain operating conditions.
The following figure from: "Low energy electron beam lithography: Monte Carlo simulation and experiment," by T. J. Stark, Z. J. Radzimski, P. A. Peterson, D. P. Griffis and P. E. Russel, in Proc. 50th Annual Meet. Electron Micros. Soc. of America, 1992
shows the effect of beam voltage on depth dose function for 50 nm tick layer of PMMA on top of a silicon substrate.
CASINO Ver. 1.0 Monte Carlo simulation program
