Charge Separation in Donor-C60 Complexes with Real-Time Green Functions: The Importance of Nonlocal Correlations

Emil Viñas Boström, Anders Mikkelsen, Claudio Verdozzi, Enrico Perfetto, Gianluca Stefanucci

Research output: Contribution to journalArticlepeer-review

25 Citations (SciVal)

Abstract

We use the nonequilibrium Green function (NEGF) method to perform real-time simulations of the ultrafast electron dynamics of photoexcited donor-C60 complexes modeled by a Pariser-Parr-Pople Hamiltonian. The NEGF results are compared to mean-field Hartree-Fock (HF) calculations to disentangle the role of correlations. Initial benchmarking against numerically highly accurate time-dependent density matrix renormalization group calculations verifies the accuracy of NEGF. We then find that charge-transfer (CT) excitons partially decay into charge separated (CS) states if dynamical nonlocal correlation corrections are included. This CS process occurs in ∼10 fs after photoexcitation. In contrast, the probability of exciton recombination is almost 100% in HF simulations. These results are largely unaffected by nuclear vibrations; the latter become however essential whenever level misalignment hinders the CT process. The robust nature of our findings indicates that ultrafast CS driven by correlation-induced decoherence may occur in many organic nanoscale systems, but it will only be correctly predicted by theoretical treatments that include time-nonlocal correlations.

Original languageEnglish
Pages (from-to)785-792
Number of pages8
JournalNano Letters
Volume18
Issue number2
DOIs
Publication statusPublished - 2018 Feb 14

Subject classification (UKÄ)

  • Condensed Matter Physics
  • Nano Technology

Keywords

  • Charge transfer
  • donor-acceptor complex
  • nonequilibrium Green functions
  • real-time simulations
  • ultrafast dynamics

Fingerprint

Dive into the research topics of 'Charge Separation in Donor-C60 Complexes with Real-Time Green Functions: The Importance of Nonlocal Correlations'. Together they form a unique fingerprint.

Cite this