A TD-DFT basis set and density functional assessment for the calculation of electronic excitation energies of fluorene

• Published in: International journal of quantum chemistry Vol. 112; no. 20; pp. 3434 - 3438
• Main Authors: Barboza, Cristina Aparecida, Vazquez, Pedro Antonio Muniz, Mac-Leod Carey, Desmond, Arratia-Perez, Ramiro
• Format: Journal Article
• Language: English
• Published: Hoboken Wiley Subscription Services, Inc., A Wiley Company 15-10-2012
• Subjects: dendrogram; excitation energies; fluorene; LR-CCSD; principal component analysis; TD-DFT
• ISSN: 0020-7608; 1097-461X
• DOI: 10.1002/qua.24300

Conjugated organic materials are the subject of intensive research for a range of optoelectronic applications. A model for such molecules is fluorene, which consists of rigid planar biphenyl units of C2v symmetry. A low energy experimental absorption spectrum in the gas phase is composed of A1 and B2 transitions. The aim of this work is to evaluate the performance of the basis sets cc‐pVXZ (X = D and T), aug‐cc‐pVDZ, 6‐31G**, 6‐31++G**, 6‐311G**, 6‐311++G**, Sadlej‐pVTZ, Z2Pol, Z3Pol, and pSBKJC and of the functionals B3LYP, B3LYP/CS00, CAM‐B3LYP, PBE0, and LB94 in predicting the electronic transitions obtained taking linear response‐coupled cluster singles and doubles (LR‐CCSD) results as the theoretical reference. Our findings suggest that the time‐dependent density functional theory singles method is not able to correctly assign the predicted spectrum while LR‐CCSD always correctly describes the experimental data. Among the studied density functionals, the best performance was achieved by the CAMB3LYP. For transitions above 5 eV, diffuse functions are required to properly predict the observed transitions. © 2012 Wiley Periodicals, Inc.
Fluorinated‐conjugated organic molecules have been the subject of intensive optoelectronic research due to their advantageous photonic and electronic properties. This study of different density functionals attempts to compare the sequence of excited states obtained with the result of a linear response‐coupled cluster singles and doubles method. Although none of the studied functionals accomplishes this, it is found that asymptotically and long‐range corrected B3LYP improved the results significantly. Finally, pseudopotentials can provide further computational cost savings.