Datafiles for sputtering yields as a function of
impact energy and angle.    

Available projectiles:
H, D, T, He, N, Ne, Ar
(only for self-sputtering): Be, C, W

Available targets:
Be, C, W

The tables are calculated with the Monte Carlo program
trvmcmom, which is vectorized version of trim.sp
determining also moments of distributions. The program
assumes a nearly flat surface, a statistically roughness
in the order of a monolayer is taken into account.
Well polished surfaces give a good agreement with the
calculated yields, but for rough surfaces the strong
angular dependence is reduced, leading to an increase
in the sputtering yield at normal incidence and a 
decrease around the sputtering maximum.

The files named [projectile]_[target].y contain the
calculated sputtering yield Y, defined as the mean
number of sputtered target atoms per unity of incident
projectile ions.

The files named [projectile]_[target].ye contain the
calculated sputtering energy yield Ye, defined in such
a way that the mean energy of sputtered atoms, for a
constant energy E0 of the incident ions, is
<E> = E0 * (Ye/Y).

The tables are arranged in such a way that rows give an
angular dependence of sputtering yields at a fixed energy
E0, and columns give an energy dependence at a fixed angle
of incidence. On top of the tables the values for the 
atomic numbers (Z) and the masses (m) of the projectile 
(index 1) and the target atoms (index 2) and the value for
the surface binding energy (Es) are given. Furthermore, the
number ne of incident energies and the number na of incident
angles at which the data are calculated are provided.

The files named [projectile]_[target]_fit.y contain the
coefficients derived fitting the calculated sputtering
yield data with the Bohdansky formula
Y(E0) = Q*Sn*(1-(Eth/E0)^(2/3))*(1-(Eth/E0))^2
where E0 is the projectile energy, and Q and Eth are
fitting parameters, contained in the tables for each
available value of the projectile incidence angle.
Sn is the nuclear stopping cross section, defined as
Sn = (0.5*log(1+1.2288*epsilon))/
     (epsilon+0.1728*sqrt(epsilon)+0.008*epsilon^0.1504)
epsilon is the reduced energy, defined as epsilon = E0/ETF,
where ETF = ((Z1*Z2*e^2)/(alpha_L))*((m1+m2)/(m2)) is the
Thomas-Fermi energy, in which Z1, Z2 are the nuclear charges,
and m1, m2 are the masses of the projectile and target atoms
respectively, and e is the electron charge. Finally, alpha_L,
defined as alpha_L = 0.4685*(Z1^(2/3)+Z2^(2/3))^(-1/2), in
angstrom, is the Lindhard screening length.

The original data are published in W. Eckstein, IPP-Report
9/117, Garching, March 1998.
The fits are produced by A. Zito, Garching, October 2022.
Status October 2022: many fits are actually not produced yet
(i.e. the coefficients are NaNs in the fit tables).