Comparisons are made between two solutions for light scattering and absorption by arbitrarily shaped or agglomerated particles. Both solutions have made improvements that result in better accuracy than that achieved by the original coupled-dipole solution of Purcell and Pennypacker [(PP); Astrophys. J.186, 705 ( 1973)]. One solution, which has an additional self-interaction term, was derived by Iskander, Chen, and Penner [(ICP); Appl. Opt.28, 3083 ( 1989)]. The other is Dungey and Bohren’s refinement of PP [(PP–DB); J. Opt. Soc. Am.8, 81 ( 1991)], which calculates the dipole polarizability from the Doyle expression instead of from the Clausius-Mossotti (CM) relation. It is found that ICP has better overall accuracy than PP–DB, especially for absorption, regarding which PP–DB is even less accurate than PP. ICP is all-around theoretically sound, whereas PP–DB is not. For nonabsorbing materials, although corrected for the violation of zero extinction for single dipoles, PP–DB creates another violation of nonzero absorption. For the dipole refractive index, the Maxwell–Garnett (MG) relation is expected to be an accurate model, since it can be approximately deduced from ICP. MG can also be used as an accurate effective refractive-index model for inhomogeneous particles.
© 1993 Optical Society of AmericaPDF Article