Your choice of software depends on what soot aggregate structure you want.
If you want a cluster of spheres, all of the same composition, it is easiest to simply use the SPHERES_N target option. You simply need a file specifying the position of the center of each sphere, and its radius (in arbitrary units). The spheres can be overlapping if desired. See the DDSCAT UserGuide for details, and the example provided in examples_exp/SPHERES_N
If you want samples of random aggregates produced by ballistic agglomeration of equal-sized spheres, see the target files available for clusters of 8, 16, 32,64, 128, 256, … spheres at
http://www.astro.princeton.edu/~draine/agglom.html
Examples are given for the standard ballistic agglomeration, and two variants (BAM1 and BAM2) that give higher density aggregates, as described by Shen, Draine & Johnson (2008: Astrophys. J., 689, 260). However, you will need to modify these files for compatibility with DDSCAT:
- delete the first line in the file
- delete the numbers in the column labeled "j"
- change all the numbers in the column labeled "2*a(j)" from 1 to 0.5
If you want a cluster of spheres of various different materials, see target option SPH_ANI_N
The value of the interdipole separation "d" is determined automatically by DDSCAT after you specify the value of AEFF = radius of equal volume sphere (in physical units, e.g., microns) and (if you are using target option SPHERES_N) the parameter SHPAR1 setting the target extent in the x direction, in units of d.
DDSCAT will report the derived value of d in physical units.