The atomic radii were taken from

J.C. Slater, J. Chem. Phys. 1964, 39, 3199.

and multiplied by 1.05 (for additional insurance that the atoms are not too close).
