Intramolecular Singlet Fission Through a Coherently Coupled Excimer-like Intermediate

TL;DR

This work tackles how electronic and nuclear motions govern intramolecular singlet fission ($iSF$) in a rigid, contorted NDI dimer. It combines polarization-controlled 2DES and impulsive pump–probe spectroscopy to resolve a coherently coupled intermediate, $[S_1+TT_{1}]$, that forms in ~$200~\mathrm{fs}$ and relaxes to $TT_{1}$ with a rate that depends on the excited Davydov component. The findings reveal an excimer-like intermediate with weak CT character, enhanced inter-chromophore vibrational beats, and surprisingly minimal electronic reorientation throughout $S_1$–$[S_1+TT_{1}]$–$TT_{1}$ evolution, implying strong electronic correlations and significant singlet–triplet mixing maintained during $TT_{1}$ formation. These results challenge current SF models, inform design rules for long-lived high-spin triplets, and demonstrate polarization–anisotropy as a powerful probe of excited-state reorientation in iSF systems.

Abstract

Singlet Fission (SF) into two triplets offers exciting avenues for high-efficiency photovoltaics and optically initializable qubits. While the chemical space of SF chromophores is ever-expanding, the mechanistic details of electronic-nuclear motions that dictate the photophysics are unclear. Rigid SF dimers with well-defined orientations are necessary to decipher such details. Here, using polarization-controlled white-light two-dimensional and pump-probe spectroscopies, we investigate a new class of contorted naphthalenediimide dimers, recently reported to have a favorable intramolecular SF (iSF) pathway. 2D cross-peaks directly identify the two Davydov components of the dimer along with strongly wavelength-dependent TT1 formation kinetics depending on which Davydov component is excited, implicating a coherently coupled intermediate that mediates iSF. Enhanced quantum beats in the TT1 photoproduct suggest that inter-chromophore twisting and ruffling motions drive the ~200 fs evolution towards an excimer-like intermediate and its subsequent ~2 ps relaxation to the TT1 photoproduct. Polarization anisotropy directly tracks electronic motion during these steps and reveals surprisingly minimal electronic reorientation with significant singlet-triplet mixing throughout the nuclear evolution away from the Franck-Condon geometry towards relaxed TT1. The observations of coherent excimer-like intermediate and significant singlet-triplet mixing throughout the iSF process need to be carefully accounted for in the synthetic design and electronic structure models for iSF dimers aiming for long-lived high-spin correlated triplets.

Figures (5)

• Figure 1: Linear excitation-emission maps reveal low-oscillator strength emissive states. (a) Linear absorption spectrum of NDI dimer in a 1:1(v/v) mixture of solvents acetonitrile and dichloromethane (ACN and DCM, respectively). The spectra are overlaid with the pump and probe white-light laser spectrum used for the impulsive PP and 2DES experiments. The NDI dimer molecule is shown on the right. (b) Emission spectrum collected as a function of excitation wavelength for dimer (top) and monomer (bottom). The 1D spectrum on the top inset shows the absorption spectrum of the dimer and the side panel shows the respective integrated emission spectrum. (c) Excitation spectrum as a function of emission collection wavelength. The top panel is the integrated emission spectrum of the dimer while the side panel is the respective absorption spectrum.
• Figure 2: 2DES reveals Davydov splitting and excitation wavelength dependent $TT_1$ formation. (a) 2D spectrum of the NDI dimer at $T$ = 0.2 ps. The top panel shows the absorption spectrum (black) overlaid with the pump spectrum (gray shaded). The side panel shows the spectrally integrated spectrum versus the detection wavelength, $\lambda_t$, overlaid with the probe spectrum (gray shaded). Contours are drawn at 10% to 100% in 10% intervals. The positive ground state bleach (GSB) signal is shown in red while the negative excited state absorption (ESA) signal is shown in blue. (b) (left) Dynamics in the ESA region denoted by the green rectangle in panel A. (middle) Dynamics in the ESA region denoted by the yellow rectangle in panel A. (right) Dynamics in the ESA region denoted by dashed yellow rectangle in panel A. For panel b (left and middle), the data was collected until 5 ps and the spectrally integrated 2D signals are fitted to a freely floated 3-exponential model where the first two time-constants are shown in the figure. For panel b (right), the time constants of the fit function for the middle panel are kept fixed and only the amplitudes are floated. The accompanying top panel shows the respective residuals. The $\pm {\sigma}$ error band on the fit is calculated by averaging N=6 trials of $T$ = 0.5 ps 2D spectrum where collection of individual trials was interleaved over the duration of the full scan.The arrows denote the 200 fs decay or rise, that is followed by the slower rise of the ESA signal. (c) 2D rate maps obtained by global fitting of all 2D pixels to a 3-exponential model. The obtained time constants are shown above the respective maps. The rate maps are normalized by the sum of the maxima of two rate maps. Global rate analysis neglects the excitation wavelength dependence seen in panel b (left versus middle). Contours are drawn from 10% to 100% in 10% intervals.
• Figure 3: Pump-probe target analysis for the NDI dimer confirms ultrafast formation of the intermediate state. (a) Spectrally-resolved PP spectrum for the NDI dimer. The marked blue and red regions denote the ESA bands in which $\sim$200 fs rise and decay of the signal is observed along the detection wavelength axis in the 2D rate maps in Figure \ref{['fig:fig2']}c. (b) Spectrally integrated pump-probe decay traces along $T$ in the blue and red ESA bands. The respective fits from the global analysis of the PP data are shown as dashed lines. (c) Species associated spectra (SAS) obtained after target analysis of the PP data. The corresponding concentration profiles of the associated species are shown in (d) for a sequential target kinetic model: $S_{1}$-$[I]$-$TT_{1}$. The integrated instrument response function (IRF) obtained from the global fit is overlaid as the grey shaded area. The gray band marks the $T$ time point where the integrated IRF is down to 1%. Table S1 (rightmost column) summarizes the pump-probe time constants obtained after the global fits. (e) Summary of all the features in the 2DES rate maps ( Figure \ref{['fig:fig2']}c) and the pump-probe data that are consistent with the target model in panels c,d.
• Figure 4: Electronic reorientation during $S_{1}$-$[S_1+TT_1]$-$TT_{1}$ internal conversion. (a) Parallel (PA) and magic angle (MA) pump-probe transients in a 10 nm GSB/SE band centered at 600 nm and ESA band centered at 725 nm. Fits from the global analysis of MA data in Figure \ref{['fig:fig3']} are overlaid starting from 100 fs. The PA fit was obtained by freely floating the same three-exponential fit function. (b) Polarization anisotropy reconstructed using the respective PA and MA data of the GSB/SE (left) and ESA (right) bands. The anisotropy derived from PA and MA transient fits are also overlaid. The anisotropy of oxazine 720 laser dye is overlaid as a reference (gray trace) in the left figure. Inset shows the zoomed in anisotropy until 1 ps. Horizontal lines in GSB/SE anisotropy mark the average anisotropy of 0.215$\pm$0.008 observed over the $T$ range of 0.12-0.15 ps. The dashed horizontal line in ESA anisotropy at 0.145$\pm$0.003 marks the average ESA anistropy observed over the same $T$ range.
• Figure 5: Enhanced quantum beats in the $TT_1$ photoproduct correspond to inter-chromophore nuclear motions. (a) Quantum beat spectrum as a function of detection wavelength. The top panel shows the PP spectrum at $T$ = 1 ps, while the side panel shows the ground state Raman spectrum of the NDI dimer excited at 785 nm along with the non-resonant Raman spectra of the two solvents ACN and DCM. The lowest contour is at 20%. Quantum beats in the PP data arising from the non-resonant Raman scattering from the solvent are marked as the horizontal grey band. (b) The vertical gray bands in panel a are spectrally integrated and overlaid for the GSB/SE (blue) and ESA (red) bands. Vertical gray bands mark the solvent bands, same as horizontal bands in panel a. (c) Schematic for the excimer-like intermediate formation due to large excited state nuclear displacements along the theoretical 219 cm$^{-1}$ and 590 cm$^{-1}$ modes that correspond to inter-chromophore ruffling and twisting motions, respectively (Section S1,S3).