Theoretical study of M(PH3)2 complexes of C60, corannulene (C20H10), and sumanene (C21H12) (M = Pd or Pt). Unexpectedly large binding energy of M(PH3)2(C60)

DFT and MP2 to MP4(SDQ) methods were applied to M(PH3)2(C60), Pt(PH3)2(C20H10), and Pt(PH3)2(C21H12) (M = Pd or Pt, C20H10 = corannulene, and C21H12 = sumanene). The binding energy considerably fluctuates around MP2 and MP3 levels but much less upon going from MP3 to MP4(SDQ) in Pt(PH3)2(C2H4), Pt(PH3)2(C20H10), and Pt(PH3)2(C21H12). Also, the MP4(SDQ) method presents a binding energy similar to that of the CCSD(T) method in Pt(PH3)2(C2H4). Thus, it is likely that the MP4(SDQ) method is useful to evaluate binding energies of these complexes. The binding energies of Pt(PH3)2(C20H10) and Pt(PH3)2(C21H12) are evaluated to be 24.9 and 26.1 kcal/mol, respectively, by the MP4(SDQ) method and only +5.8 and -2.6 kcal/mol, respectively, by the DFT(B3LYP) method. These MP4(SDQ)-calculated binding energies of Pt(PH3)2(C20H10) and Pt(PH3)2(C21H12) are similar to that of Pt(PH3)2(C2H4), which strongly suggests that these complexes can be successfully synthesized. The binding energy of Pt(PH3)2(C60) is evaluated to be 44.8 and 45.5 kcal/mol with the ONIOM(MP4(SDQ):UFF) and ONIOM(MP4(SDQ):B3LYP) methods, respectively, and that of the Pd analogue is evaluated to be 39.9 kcal/mol with the ONIOM(MP4(SDQ):UFF) method, whereas the DFT(B3LYP), DFT(BVP86), and DFT(BPW91) methods provide much smaller binding energies. It is noted that these binding energies are much larger than those of the ethylene, corannulene, and sumanene analogues. This difference is reasonably interpreted in terms that the LUMO of C60 is at much lower energy than those of ethylene, corannulene, and sumanene. We investigated also how to separate the high level and the low level regions in the ONIOM calculation of M(PH3)2(C60) and proposed here the reasonable way to evaluate the binding energy of transition-metal complexes of C60.