The most important issue is in obtaining an appropriate geometry for your simulation. Evaluating the absorption spectrum is relatively straight forward with, for example, TDDFT on the optimal ground state structure, but the emission spectrum requires knowledge of excited state geometries, which are expensive and challenging to simulate. Do you know how large the Stokes shift in your system is? It may be large if your nanocluster is relatively flexible and excited state geometries would then be a necessity. I don't know if Gaussian is capable of performing efficient excited state geometry optimisations. if it is not, you could consider Turbomole, which has this capability.
Thanks Mr. Andrew, yeah I've noticed this, not only geometry, I'm trying to focus on solvent media, ligand chemistry , shape and size of NPs/NCs. but Its harder to find a good software, a number of softwares are available, I just started, I'm evaluating the reasonable and good software, may be series of softwares will be good, as no single is perfect for calculations.