A Density Functional Theory Study of Porphyrin-Pyridine-Fullerene Triad ZnTPP•Py•C60

We report on DFT calculations of noncovalent porphyrin-pyridine-fullerene triad ZnTPP center dot Py center dot C-60 (ZnTPP = zinc(II) complex of meso-tetraphenylporphine; Py = pyridine), using PW91, PBE and BLYP functionals of general gradient approximation (GGA), as well as PWC and VWN functionals of local density approximation (LDA) implemented in the DMol3 module of Materials Studio. The optimized geometries were compared to the corresponding experimental parameters obtained previously by X-ray diffraction study of ZnTPP center dot BPy center dot C-60 complex (BPy = bipyridyl). If the correlation coefficient between the calculated and experimental X-ray geometric parameters is used, all DFT functionals tested show a very similar performance for the covalent/coordination bond lengths and angles within ZnTPP center dot Py unit. If the root mean square error is applied, BLYP is the best functional for calculating bond angles and worst for calculating bond lengths, and vice versa for LDA functionals. The separation between ZnTPP and fullerene units, expressed as N(ZnTPP)center dot center dot center dot C(C-60) and Zn center dot center dot center dot C(C-60) distances, was best reproduced using PWC and VWN LDA functionals (slightly shortened by 0.1-0.2 angstrom). PW91 and PBE GGA functionals overestimated the experimental N(ZnTPP)center dot center dot center dot C(C-60) and Zn center dot center dot center dot C(C-60) distances by roughly 0.4 angstrom. The worst results were obtained by BLYP, which produced totally unrealistic separations. Taken together with positive formation energy obtained with BLYP (which implies repulsive interaction), the results obtained suggest that PWC and VWN are the most appropriate, whereas BLYP is the least recommendable functional of DMol3 for the studies of noncovalent interactions of porphyrins with fullerenes and carbon nanotubes.